正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

#compchem Second preprint linked to the FeNNix-Bio1 #machinelearning foundation model. FeNNix-Bio1's inference is pretty fast already with a few GPUs but, "what if", we were able to push it at the #Exascale? Let's have a glimpse into the future (1/3): "Pushing the Accuracy Limit of Foundation Neural Network Models... with Quantum Monte Carlo Forces and Path Integrals" Check it out 💫: We propose an end-to-end integrated strategy to produce highly accurate quantum chemistry synthetic datasets (energies and forces) aimed at deriving Foundation Machine Learning models for molecular simulation. Starting from Density Functional Theory (DFT), a "Jacob's Ladder" approach leverages computationally-optimized layers of massively #GPU-accelerated software with increasing accuracy. Thanks to Exascale, this is the first time that the computationally intensive calculation of Quantum Monte Carlo forces (QMC), and the combination of multi-determinant QMC energies and forces with selected-Configuration Interaction wavefunctions, are computed at such scale at the complete basis-set limit. To bridge the gap between accurate QC and condensed-phase Molecular Dynamics, we leverage transfer learning to improve the DFT-based FeNNix-Bio1 foundation model. 🚀The resulting approach is coupled to path integrals adaptive sampling quantum dynamics to perform nanosecond reactive simulations at unprecedented accuracy on a full Satellite Tobacco Mosaic Virus (STMV) 1M, all-atom, complete solvated model (see the video produced using VTX, Maxime MARIA Matthieu Montes ). These results demonstrate the promise of Exascale to deepen our understanding of the inner machinery of complex biosystems. Immense thanks to all co-authors at Qubit Pharmaceuticals, Laboratoire de Chimie Théorique (Sorbonne Université /CNRS 🌍 ), The University of Chicago, Sandia National Labs, Oak Ridge Lab and Argonne National Lab for this collaborative efforts. Some are on X: Anouar Benali Thomas Plé ADJOUA Olivier Evgeny Posenitskiy Margaret Blazhynska Thomas Applencourt Jeongnim Kim #HPC This work was made possible thanks to #INCITE projects enabling the use of Argonne's Aurora exascale system and of the Polaris machine, to Genci (Jean Zay @ Idris) and EuroHPC Joint Undertaking (Leonardo CINECA). #supercomputingshow more

Jean-Philip Piquemal

7,183 subscribers

15,150 次观看 • 1 年前 •via X (Twitter)

科学技术 #compchem #machinelearning #Exascale

Anya Rossi• Live Now

Private livecam show

7 条评论

Jean-Philip Piquemal 的头像

Jean-Philip Piquemal1 年前

(2/3) We provided a new #Exascale implementation of QMCPACK enabling the efficient computation of Diffusion Monte Carlo energies and forces. Single determinant and multideterminant forces are available. We also proposed the integration of selected Configuration (sCI) into the QMC force framework to directly import the sCI multideterminental wavefunctions into QMCPACK to enable sCI computations of both energies and forces.

Rainmaker 的头像

Rainmaker2 年前

Can Machine Learning beat the market? Check out this post on my free Substack where I share code and commentary for an XGBoost model and a Random Forest model that both deliver powerful performances.

Jean-Philip Piquemal 的头像

Jean-Philip Piquemal1 年前

(3/3) To bridge the gap between accurate quantum chemistry and condensed-phase Molecular Dynamics, we leverage transfer learning to improve the DFT-based FeNNix-Bio1 foundation model using the DMC/sCI energies and forces. The resulting approach is coupled to path integrals adaptive sampling quantum dynamics to perform nanosecond reactive simulations at unprecedented accuracy. We used this "beyond DFT" model to study the PH transition (PH=7 to 5) of a full, 1M-atom Sattelite Tobacco Virus model.

Jean-Philip Piquemal 的头像

Jean-Philip Piquemal1 年前

Nice video by @blazhynska66497

Jean-Philip Piquemal 的头像

Jean-Philip Piquemal1 年前

To know more about the FeNNix-Bio1 model, check the preprint:

Antonio Alvarez de la Paz 的头像

Antonio Alvarez de la Paz1 年前

Impressive, congratulations!

Richard Collins, The Internet Foundation 的头像

Richard Collins, The Internet Foundation1 年前

My comment. ( I actually meant the calculation should not rotate, but Grok's take works too.)

相关视频

D-Wave announced a scientific breakthrough published in the esteemed journal Science Magazine, confirming that its annealing quantum computer outperformed one of the world’s most powerful classical supercomputers in solving a complex magnetic materials simulation problem with relevance to materials discovery. The new landmark peer-reviewed paper, “Beyond-Classical Computation in Quantum Simulation,” validates this achievement as the world’s first and only demonstration of quantum computational supremacy on a useful problem. An international collaboration of scientists led by D-Wave performed simulations of quantum dynamics in programmable spin glasses—a computationally hard magnetic materials simulation problem with known applications to business and science—on both D-Wave’s Advantage2™ prototype annealing quantum computer and the Frontier supercomputer at the Department of Energy’s Oak Ridge Lab. D-Wave’s quantum computer performed a complex simulation in minutes and with a level of accuracy that would take nearly a million years using the supercomputer. In addition, it would require more than the world’s annual electricity consumption to solve this problem using the supercomputer, which is built with graphics processing unit (GPU) clusters. For decades, scientists have aspired to build a quantum computer capable of solving complex materials simulation problems beyond the reach of classical computers. D-Wave's advancements in quantum hardware have made it possible for its annealing quantum computers to process these types of problems for the first time. Magnetic materials simulations, like those conducted in this work, use computer models to study how tiny particles not visible to the human eye react to external factors. Magnetic materials are widely used in medical imaging, electronics, superconductors, electrical networks, sensors, and motors. This is an incredibly important achievement. Please join us in congratulating the D-Wave team and our global collaborators on this remarkable milestone. It’s a significant moment for the quantum computing industry. Learn more about this monumental achievement: Read the press release here: #QuantumSupremacy #QuantumRealized #QuantumComputing #DWave #Technology #Innovation #Optimization #MaterialsDiscovery #ScientificBreakthrough $QBTS

D-Wave announced a scientific breakthrough published in the esteemed journal Science Magazine, confirming that its annealing quantum computer outperformed one of the world’s most powerful classical supercomputers in solving a complex magnetic materials simulation problem with relevance to materials discovery. The new landmark peer-reviewed paper, “Beyond-Classical Computation in Quantum Simulation,” validates this achievement as the world’s first and only demonstration of quantum computational supremacy on a useful problem. An international collaboration of scientists led by D-Wave performed simulations of quantum dynamics in programmable spin glasses—a computationally hard magnetic materials simulation problem with known applications to business and science—on both D-Wave’s Advantage2™ prototype annealing quantum computer and the Frontier supercomputer at the Department of Energy’s Oak Ridge Lab. D-Wave’s quantum computer performed a complex simulation in minutes and with a level of accuracy that would take nearly a million years using the supercomputer. In addition, it would require more than the world’s annual electricity consumption to solve this problem using the supercomputer, which is built with graphics processing unit (GPU) clusters. For decades, scientists have aspired to build a quantum computer capable of solving complex materials simulation problems beyond the reach of classical computers. D-Wave's advancements in quantum hardware have made it possible for its annealing quantum computers to process these types of problems for the first time. Magnetic materials simulations, like those conducted in this work, use computer models to study how tiny particles not visible to the human eye react to external factors. Magnetic materials are widely used in medical imaging, electronics, superconductors, electrical networks, sensors, and motors. This is an incredibly important achievement. Please join us in congratulating the D-Wave team and our global collaborators on this remarkable milestone. It’s a significant moment for the quantum computing industry. Learn more about this monumental achievement: Read the press release here: #QuantumSupremacy #QuantumRealized #QuantumComputing #DWave #Technology #Innovation #Optimization #MaterialsDiscovery #ScientificBreakthrough $QBTS

D-Wave

65,039 次观看 • 1 年前

This can be the superpower of the Neural Band that Meta is giving together with the Ray-Ban Meta Display glasses. The video shows an old prototype bracelet by CTRL+LABS, the startup acquired by Meta and whose technology was used to develop the Neural Band. At the beginning of the video, the guy makes an action (a keyboard key press) with his hands, then the bracelet can substitute the key pressure, and at the end of the video, the guy doesn't even have to do the action; it is just sufficient that he "thinks" about it. As long as the brain is sending an electric message to the fingers, the full action is not necessary anymore. Just an "intention" to move them is necessary. If the Neural Band is evolved to this stage, and the users are educated to this, potentially, we may not even need to perform air taps or writing gestures, but we could just think about doing them. This would reduce a lot of the fatigue of using XR devices and the weirdness of using them on the street. Then why isn't this feature available today? I guess that the reason is twofold. First of all, we have accuracy: the full gesture is easier to detect for the system. Many people (me included) are praising the accuracy of the Neural Band, and this is amazing, because an input mechanism should have a reliability close to 100%. Then we, as users, have never been trained to just "think" about actions: it would feel weird and hard to learn. I think we should undergo some training to learn how to do this "thinking" operation properly. I hope that something like this could come in the upcoming years... that would be the real game-changer paradigm if compared to the camera-based tracking.

This can be the superpower of the Neural Band that Meta is giving together with the Ray-Ban Meta Display glasses. The video shows an old prototype bracelet by CTRL+LABS, the startup acquired by Meta and whose technology was used to develop the Neural Band. At the beginning of the video, the guy makes an action (a keyboard key press) with his hands, then the bracelet can substitute the key pressure, and at the end of the video, the guy doesn't even have to do the action; it is just sufficient that he "thinks" about it. As long as the brain is sending an electric message to the fingers, the full action is not necessary anymore. Just an "intention" to move them is necessary. If the Neural Band is evolved to this stage, and the users are educated to this, potentially, we may not even need to perform air taps or writing gestures, but we could just think about doing them. This would reduce a lot of the fatigue of using XR devices and the weirdness of using them on the street. Then why isn't this feature available today? I guess that the reason is twofold. First of all, we have accuracy: the full gesture is easier to detect for the system. Many people (me included) are praising the accuracy of the Neural Band, and this is amazing, because an input mechanism should have a reliability close to 100%. Then we, as users, have never been trained to just "think" about actions: it would feel weird and hard to learn. I think we should undergo some training to learn how to do this "thinking" operation properly. I hope that something like this could come in the upcoming years... that would be the real game-changer paradigm if compared to the camera-based tracking.

TonyVT SkarredGhost

11,805 次观看 • 10 个月前

6,100-Qubit Processor Shatters Quantum Computing Record | David Nield, ScienceAlert Another major quantum computing record has been broken, and by a considerable margin: physicists have now built an array containing 6,100 qubits, the largest of its type and way above the thousand or so qubits previous systems contained. It's the work of scientists from the California Institute of Technology, who used cesium atoms as their qubits, trapping them in place with a complex system of lasers that acted as tweezers to keep the atoms as stable as possible. Qubits differ from the classical bits of traditional computers by exploiting what's known as a superposition: not just binary states of 1 or 0, but a spread of probabilities that allows for algorithms that can solve problems considered out of reach of conventional computing methods. Related: Quantum Advantage: A Physicist Explains The Future of Computers A lot of qubits will be needed to make quantum algorithms practical, however. One reason for these large arrays is error correction, which helps overcome the inherent fragility of the qubit by providing a surplus to double-check the machine's operation. "This is an exciting moment for neutral-atom quantum computing," says physicist Manuel Endres. "We can now see a pathway to large error-corrected quantum computers. The building blocks are in place." There was no single breakthrough that enabled this jump in qubit numbers, but rather a series of engineering advancements in many key areas – from the laser tweezers to the ultra-high (very low pressure) vacuum chamber. Stability has also been a problem for quantum computing systems. The innovations in this latest array kept qubits in a superposition state for almost 13 seconds – almost ten times longer than previous configurations had managed. What's more, individual qubits could be manipulated with 99.98 percent accuracy, establishing a significant benchmark in the programmability of quantum technology. "Large scale, with more atoms, is often thought to come at the expense of accuracy, but our results show that we can do both," says physicist Gyohei Nomura. "Qubits aren't useful without quality. Now we have quantity and quality." To make quantum computers a practical alternative to modern supercomputers, more qubits and even greater levels of stability will be required. Experts are tackling the problem from several different angles, which is why records for some types of quantum computer don't necessarily apply to others. Next, the researchers need to work on exploiting entanglement, which will enable the system to make the leap from storing information to actually processing it. Not too far in the future, we could be using these computers to discover new materials, matter, and fundamental laws of physics. "It's exciting that we are creating machines to help us learn about the Universe in ways that only quantum mechanics can teach us," says physicist Hannah Manetsch. Read more:

6,100-Qubit Processor Shatters Quantum Computing Record | David Nield, ScienceAlert Another major quantum computing record has been broken, and by a considerable margin: physicists have now built an array containing 6,100 qubits, the largest of its type and way above the thousand or so qubits previous systems contained. It's the work of scientists from the California Institute of Technology, who used cesium atoms as their qubits, trapping them in place with a complex system of lasers that acted as tweezers to keep the atoms as stable as possible. Qubits differ from the classical bits of traditional computers by exploiting what's known as a superposition: not just binary states of 1 or 0, but a spread of probabilities that allows for algorithms that can solve problems considered out of reach of conventional computing methods. Related: Quantum Advantage: A Physicist Explains The Future of Computers A lot of qubits will be needed to make quantum algorithms practical, however. One reason for these large arrays is error correction, which helps overcome the inherent fragility of the qubit by providing a surplus to double-check the machine's operation. "This is an exciting moment for neutral-atom quantum computing," says physicist Manuel Endres. "We can now see a pathway to large error-corrected quantum computers. The building blocks are in place." There was no single breakthrough that enabled this jump in qubit numbers, but rather a series of engineering advancements in many key areas – from the laser tweezers to the ultra-high (very low pressure) vacuum chamber. Stability has also been a problem for quantum computing systems. The innovations in this latest array kept qubits in a superposition state for almost 13 seconds – almost ten times longer than previous configurations had managed. What's more, individual qubits could be manipulated with 99.98 percent accuracy, establishing a significant benchmark in the programmability of quantum technology. "Large scale, with more atoms, is often thought to come at the expense of accuracy, but our results show that we can do both," says physicist Gyohei Nomura. "Qubits aren't useful without quality. Now we have quantity and quality." To make quantum computers a practical alternative to modern supercomputers, more qubits and even greater levels of stability will be required. Experts are tackling the problem from several different angles, which is why records for some types of quantum computer don't necessarily apply to others. Next, the researchers need to work on exploiting entanglement, which will enable the system to make the leap from storing information to actually processing it. Not too far in the future, we could be using these computers to discover new materials, matter, and fundamental laws of physics. "It's exciting that we are creating machines to help us learn about the Universe in ways that only quantum mechanics can teach us," says physicist Hannah Manetsch. Read more:

Owen Gregorian

43,078 次观看 • 10 个月前

Statement by Joint Force of the Armed Movement Joint Forces of the Armed Struggle Movements Military Statement To the proud masses of our people, In light of the repeated conspiracies targeting the security and safety of our homeland, the intelligence of the Joint Forces of Armed Struggle Movements, in coordination with the Sudanese Military Intelligence Authority , has been monitoring hostile activities aimed at smuggling unprecedented dangerous weapons into Sudan. These weapons were intended to be transported from a European country through Chad to Darfur, specifically to the city of El Fasher, with the goal of seizing the city using devastating arms. To the proud masses of our people, Thanks to precise intelligence monitoring from both across borders and within the country, the Special Operations Forces of the Joint Forces, in coordination with the Sudanese Army, succeeded in thwarting the delivery of these dangerous weapons. A meticulously planned military operation was carried out in the X area near Nyala at 3:00 AM on Sunday, December 1, 2024. Outcomes of the operation: •Seizure of 3 large drones capable of carrying four air-to-surface missiles each. •Seizure of 6 smaller drones capable of carrying two air-to-surface missiles each. •Elimination of over 30 mercenaries from the Rapid Support Forces militia. •Destruction of 8 military vehicles. Details of the seized weapons: The weapons included drones of the Danger Propellers model, bearing the serial number J/DF24-23-01-043A, manufactured by Woodcomp Propellers in the Czech Republic, Central Europe. These drones were modified for military purposes. Their entry into Sudan marks an unprecedented event, facilitated and funded by the United Arab Emirates (UAE) and coordinated with Chad to ensure their passage. Message to the UAE: You must understand that we are tracking all your steps from the western and northern borders. Your plans will fail no matter how much effort and conspiracy you invest. Cease tampering with our land and security. Know that you are wasting your people’s wealth on a doomed project, and defeat will be your fate in Sudan. We will remain steadfast, determined to eliminate all your regional and international mercenaries and seize your weapons, no matter their cost. We also thank you for your generosity in delivering these sophisticated weapons, valued at millions, which are now in our hands. We will use them to further your defeat. Message to the Chadian regime: The Sudanese and Chadian peoples are bound by ties of blood, history, and shared destiny. Your support for the Rapid Support Forces militia with weapons and supplies, and your facilitation of their passage through your territory, is an act of betrayal that history will not forgive. Your facilitation of these weapons and your support for the militia is a stain of shame, and you will pay a heavy price if you do not stop this recklessness. To the Sudanese people: This exceptional military achievement reflects the readiness of our forces to confront all conspiracies targeting the security and safety of our homeland. We renew our pledge to you to remain vigilant and steadfast, with our hands on the trigger, protecting the lives of our people and the resources of our nation from any external or internal threats until we achieve full liberation. Glory and immortality to our righteous martyrs. Shame and disgrace to the enemies of the homeland. Major Ahmed Hussein Mustafa Official Spokesperson of the Joint Forces of Armed Struggle Movements El Fasher December 1, 2024

Statement by Joint Force of the Armed Movement Joint Forces of the Armed Struggle Movements Military Statement To the proud masses of our people, In light of the repeated conspiracies targeting the security and safety of our homeland, the intelligence of the Joint Forces of Armed Struggle Movements, in coordination with the Sudanese Military Intelligence Authority , has been monitoring hostile activities aimed at smuggling unprecedented dangerous weapons into Sudan. These weapons were intended to be transported from a European country through Chad to Darfur, specifically to the city of El Fasher, with the goal of seizing the city using devastating arms. To the proud masses of our people, Thanks to precise intelligence monitoring from both across borders and within the country, the Special Operations Forces of the Joint Forces, in coordination with the Sudanese Army, succeeded in thwarting the delivery of these dangerous weapons. A meticulously planned military operation was carried out in the X area near Nyala at 3:00 AM on Sunday, December 1, 2024. Outcomes of the operation: •Seizure of 3 large drones capable of carrying four air-to-surface missiles each. •Seizure of 6 smaller drones capable of carrying two air-to-surface missiles each. •Elimination of over 30 mercenaries from the Rapid Support Forces militia. •Destruction of 8 military vehicles. Details of the seized weapons: The weapons included drones of the Danger Propellers model, bearing the serial number J/DF24-23-01-043A, manufactured by Woodcomp Propellers in the Czech Republic, Central Europe. These drones were modified for military purposes. Their entry into Sudan marks an unprecedented event, facilitated and funded by the United Arab Emirates (UAE) and coordinated with Chad to ensure their passage. Message to the UAE: You must understand that we are tracking all your steps from the western and northern borders. Your plans will fail no matter how much effort and conspiracy you invest. Cease tampering with our land and security. Know that you are wasting your people’s wealth on a doomed project, and defeat will be your fate in Sudan. We will remain steadfast, determined to eliminate all your regional and international mercenaries and seize your weapons, no matter their cost. We also thank you for your generosity in delivering these sophisticated weapons, valued at millions, which are now in our hands. We will use them to further your defeat. Message to the Chadian regime: The Sudanese and Chadian peoples are bound by ties of blood, history, and shared destiny. Your support for the Rapid Support Forces militia with weapons and supplies, and your facilitation of their passage through your territory, is an act of betrayal that history will not forgive. Your facilitation of these weapons and your support for the militia is a stain of shame, and you will pay a heavy price if you do not stop this recklessness. To the Sudanese people: This exceptional military achievement reflects the readiness of our forces to confront all conspiracies targeting the security and safety of our homeland. We renew our pledge to you to remain vigilant and steadfast, with our hands on the trigger, protecting the lives of our people and the resources of our nation from any external or internal threats until we achieve full liberation. Glory and immortality to our righteous martyrs. Shame and disgrace to the enemies of the homeland. Major Ahmed Hussein Mustafa Official Spokesperson of the Joint Forces of Armed Struggle Movements El Fasher December 1, 2024

الحساب الرسمي للقوة المشتركة لحركات الكفاح المسلح

28,991 次观看 • 1 年前

Small Language Models (SML) are the future of AI. "Small" (SML) instead of "Large" (LLM). These small models are highly specialized models with superhuman abilities on specific tasks. Here are two techniques to build these models: • Spectrum • Model Merging I give you a short introduction in the attached video, but here is a quick summary: Spectrum helps us identify the most relevant layers to solve one specific task. We can ignore everything else and focus on fine-tuning these layers. Using Spectrum, we can fine-tune models in a heartbeat. Model Merging combines multiple models into a unique, much better model than any of the individual input models. You can also combine models specialized in different tasks and get a model with multiple abilities. This is the state of the art of productizing models. It's what Arcee.ai's platform does behind the scenes. Arcee collaborated with me on this post and is sponsoring it. There are three main steps to produce a model for your particular use case: 1. You create a dataset by uploading your data. 2. You train a model. At this step, Arcee uses Spectrum and Model Merging to produce a highly specialized model for your task. 3. You can deploy that model to any environment you want. Three important notes: • Training process is 2x faster and 2x cheaper than regular fine-tuning. • Resultant models are smaller and have higher accuracy. • They create these specialized models from open-source models. Check this site so you can fully appreciate how this works: If you want to fine-tune an open-source model, consider Arcee's platform. This is the state of the art.

Small Language Models (SML) are the future of AI. "Small" (SML) instead of "Large" (LLM). These small models are highly specialized models with superhuman abilities on specific tasks. Here are two techniques to build these models: • Spectrum • Model Merging I give you a short introduction in the attached video, but here is a quick summary: Spectrum helps us identify the most relevant layers to solve one specific task. We can ignore everything else and focus on fine-tuning these layers. Using Spectrum, we can fine-tune models in a heartbeat. Model Merging combines multiple models into a unique, much better model than any of the individual input models. You can also combine models specialized in different tasks and get a model with multiple abilities. This is the state of the art of productizing models. It's what Arcee.ai's platform does behind the scenes. Arcee collaborated with me on this post and is sponsoring it. There are three main steps to produce a model for your particular use case: 1. You create a dataset by uploading your data. 2. You train a model. At this step, Arcee uses Spectrum and Model Merging to produce a highly specialized model for your task. 3. You can deploy that model to any environment you want. Three important notes: • Training process is 2x faster and 2x cheaper than regular fine-tuning. • Resultant models are smaller and have higher accuracy. • They create these specialized models from open-source models. Check this site so you can fully appreciate how this works: If you want to fine-tune an open-source model, consider Arcee's platform. This is the state of the art.

Santiago

164,162 次观看 • 2 年前

A common misconception in physics is that gravity must be negligible at the quantum scale because its measured strength appears weak compared to electromagnetism and the strong nuclear force. However, this perception arises from considering gravity only in its weak field limit. Obviously, it will appear weak under such conditions, but the situation changes when we consider regions of high mass-energy density. For example, when examining strong gravitational fields near black holes, the force becomes immensely powerful. With this in mind, lets re-examine unification and gravity at the quantum scale: Following our holographic mass solution, elementary particles like protons are found to be microscopic Schwarzschild black holes and their mass-energy density is sufficient to create a gravitational curvature equivalent to the strong nuclear force. The vacuum energy density of space provides enough energy to maintain these structures through Planck-scale dynamics, preventing their rapid evaporation via Hawking radiation, so they are ultra-stable. These discoveries, expounded in our studies like The Origin of Mass and the Nature of Gravity, available to download for free on CERN's Zenodo preprint server-🔗 -demonstrate that what we observe as distinct fundamental forces are in fact manifestations of a unified force operating at different scales, where quantum vacuum fluctuations significantly curve spacetime (the curvature being gravity) with the resulting encapsulation generating screening effects that modulate the apparent strength of gravity from its fundamental high-energy state, which results in nuclear confinement forces, to its familiar weak-field behavior that appears as regular Newtonian gravity.

A common misconception in physics is that gravity must be negligible at the quantum scale because its measured strength appears weak compared to electromagnetism and the strong nuclear force. However, this perception arises from considering gravity only in its weak field limit. Obviously, it will appear weak under such conditions, but the situation changes when we consider regions of high mass-energy density. For example, when examining strong gravitational fields near black holes, the force becomes immensely powerful. With this in mind, lets re-examine unification and gravity at the quantum scale: Following our holographic mass solution, elementary particles like protons are found to be microscopic Schwarzschild black holes and their mass-energy density is sufficient to create a gravitational curvature equivalent to the strong nuclear force. The vacuum energy density of space provides enough energy to maintain these structures through Planck-scale dynamics, preventing their rapid evaporation via Hawking radiation, so they are ultra-stable. These discoveries, expounded in our studies like The Origin of Mass and the Nature of Gravity, available to download for free on CERN's Zenodo preprint server-🔗 -demonstrate that what we observe as distinct fundamental forces are in fact manifestations of a unified force operating at different scales, where quantum vacuum fluctuations significantly curve spacetime (the curvature being gravity) with the resulting encapsulation generating screening effects that modulate the apparent strength of gravity from its fundamental high-energy state, which results in nuclear confinement forces, to its familiar weak-field behavior that appears as regular Newtonian gravity.

Nassim Haramein

19,882 次观看 • 1 年前

I'm thrilled to announce the launch of and the start of a new conversation. We can all feel it: the web is changing in ways that seem both dramatic and, in a weird way, imperceptible. At times, it’s hard to know where to start exploring what could be next. Sometimes, when you think you've found whatever next could be, it's hard to determine what’s real and what’s hype. Our approach has always been simple: We like to put a big question out there on the horizon of what’s possible, then build toward it with conviction to see what we can learn. When we created the new Edgar Allan site, we set out to explore the impact of conversational design and personalization in marketing experiences. Why conversation? Conversation and dialogue are the foundation of all stories and the vehicle for human connection. Today, it’s difficult to ignore the prevalence of conversation in the digital world, from Slack and Teams channels to social media feeds, phone messaging, and tools like ChatGPT. So, with that in mind, we had some thoughts: ‍“What if a website was less a one-sided place to present information and more of a living knowledge center?” “What if a site could be more of a guide than a brochure?” “What if we could incrementally learn about a visitor’s intent, then dynamically shorten the distance between everything a brand says and the mere seconds of attention most websites get?” Ultimately, the agency's job is to work at the edge of the future and what’s possible so we can make things and report the results to our clients and community. Thank you for joining us in this process. We can’t wait to see how this conversation unfolds and what we can learn together. Also, a huge thank you to the entire EA team. After 258 days of wild exploration and untold effort, today, we are starting the process of learning, designing, building, and optimizing. This site feels like us, and I couldn't be more excited.

Edgar Allan

28,025 次观看 • 1 年前

Introducing FLUX-mimic, a next-generation Video-Action Model for general purpose dexterity, developed in partnership with Black Forest Labs. Late last year we published mimic-video and introduced Video-Action Models (VAM): a new family of robotics foundation models built on top of video generation models. We showed that robot control reduces to visual prediction, and that robot capability is downstream of improvements in video modeling accuracy. The obvious implication was that advances in the video modeling frontier would directly translate to increased capabilities in end-to-end robot learning. FLUX-mimic is that thesis at frontier scale: We've applied our VAM architecture to the strongest video backbone available today, FLUX 3 from Black Forest Labs, and trained it on data from our own robots and wearables. General-purpose dexterity, running on a single GPU on premises. Because the model already understands world dynamics, it needs far fewer demonstrations to learn a new task. This is game-changing for our mission to deploy robots to factory floors, where industrial robot data is scarce and expensive to collect. We're now testing and deploying FLUX-mimic with manufacturing leaders like Audi USA, on complex, multi-step manipulation long considered impossible for conventional automation.

Introducing FLUX-mimic, a next-generation Video-Action Model for general purpose dexterity, developed in partnership with Black Forest Labs. Late last year we published mimic-video and introduced Video-Action Models (VAM): a new family of robotics foundation models built on top of video generation models. We showed that robot control reduces to visual prediction, and that robot capability is downstream of improvements in video modeling accuracy. The obvious implication was that advances in the video modeling frontier would directly translate to increased capabilities in end-to-end robot learning. FLUX-mimic is that thesis at frontier scale: We've applied our VAM architecture to the strongest video backbone available today, FLUX 3 from Black Forest Labs, and trained it on data from our own robots and wearables. General-purpose dexterity, running on a single GPU on premises. Because the model already understands world dynamics, it needs far fewer demonstrations to learn a new task. This is game-changing for our mission to deploy robots to factory floors, where industrial robot data is scarce and expensive to collect. We're now testing and deploying FLUX-mimic with manufacturing leaders like Audi USA, on complex, multi-step manipulation long considered impossible for conventional automation.

mimic

116,113 次观看 • 20 天前

How IIT Madras is Changing Global Security The End of Hacking? IIT Madras is spearheading India's push into quantum-secure communications through the IITM-C-DOT Samgnya Technologies Foundation, launched as the National Hub for Quantum Communication under the National Quantum Mission. This initiative aims to make communications "unhackable" by shifting from math-based encryption to physics-based security, countering future quantum computer threats. The hub, inaugurated in December 2025 at IIT Madras Research Park, partners with C-DOT and is funded by the Department of Science and Technology (DST). It focuses on developing indigenous quantum hardware, QKD networks, quantum repeaters, and satellite-based systems to protect critical infrastructure like AI data centers and defense networks. Traditional encryption relies on complex math that quantum computers will crack in years. QKD uses quantum physics: any eavesdropping alters the quantum state, alerting the system instantly and burning the key. IIT Madras claims an edge in secure transmission over computing leaders like the US and China. The hub builds real-world testbeds, subsidizes hardware for startups, trains experts, and links globally via IITM Global outposts. It positions India to lead in quantum-secure networks for sovereign AI and government use within five years. IIT Madras also runs CyStar, a cybersecurity center advancing quantum security, AI model protection, and IoT defenses since 2024. This aligns with national goals for unhackable comms amid rising cyber threats. Credit : AIM Network.

How IIT Madras is Changing Global Security The End of Hacking? IIT Madras is spearheading India's push into quantum-secure communications through the IITM-C-DOT Samgnya Technologies Foundation, launched as the National Hub for Quantum Communication under the National Quantum Mission. This initiative aims to make communications "unhackable" by shifting from math-based encryption to physics-based security, countering future quantum computer threats. The hub, inaugurated in December 2025 at IIT Madras Research Park, partners with C-DOT and is funded by the Department of Science and Technology (DST). It focuses on developing indigenous quantum hardware, QKD networks, quantum repeaters, and satellite-based systems to protect critical infrastructure like AI data centers and defense networks. Traditional encryption relies on complex math that quantum computers will crack in years. QKD uses quantum physics: any eavesdropping alters the quantum state, alerting the system instantly and burning the key. IIT Madras claims an edge in secure transmission over computing leaders like the US and China. The hub builds real-world testbeds, subsidizes hardware for startups, trains experts, and links globally via IITM Global outposts. It positions India to lead in quantum-secure networks for sovereign AI and government use within five years. IIT Madras also runs CyStar, a cybersecurity center advancing quantum security, AI model protection, and IoT defenses since 2024. This aligns with national goals for unhackable comms amid rising cyber threats. Credit : AIM Network.

Augadh

11,978 次观看 • 6 个月前

Today, I'm releasing the first eval meant to test whether frontier models will help with authoritarian requests, or resist--the Dictatorship Eval. Headline finding: while some models resist direct authoritarian requests, they all comply with requests disguised as innocuous edits to codebases. As AI is woven into the government and so many parts of society, the biggest near-term risk for freedom isn't some scifi dictatorship of a runaway AI: it's people inside government or inside model companies using the technology to suppress or control us. Model companies understand this, and several of them (particularly Anthropic and OpenAI) have written explicit policies meant to prevent the models from going along with nefarious requests like these. But how well are these policies playing out in practice? Despite all the recent discussion of these issues around the conflict between Anthropic and the Pentagon, no one has systematically tested what the models actually do in these contexts, as opposed to what people in government and industry say they're supposed to do. That's what the Dictatorship Eval does. And the findings suggest we have a lot of work to do to align the policies with what really goes on in practice. It's hard to define what counts as an authoritarian request, so I'm open sourcing the whole library of scenarios I used so that others can improve on them. It's also hard to get an accurate picture of how the models might be used for authoritarian ends, because I can only test hypothetical requests using public-facing models, while the government and the model companies can obviously use internal models with different guardrails. But hopefully this work is a useful first step that gives us some sense of what's going on, and a sort of "lower bound" on how models comply with these requests. Finally: it's not obvious to me that the correct solution here is increasing the rate at which models refuse these requests. Do we really want models scanning our code and judging its moral value before agreeing to help us? Or should we double down on improving how we govern against authoritarianism at the societal level, while leaving the tools open to fulfilling most requests? The answer is probably in between. Just like we don't want the models to help create bioweapons, we probably do want them to explicitly refuse outrageous requests. But we probably also want to limit how often and how strongly they refuse and fall back on other means for guarding against their use for authoritarian ends. I'm super grateful to everyone who gave me feedback on this project along the way, especially Ethan BdM , Zhengdong , Connor Huff, and a bunch of folks at Anthropic. Looking forward to getting feedback from the community and iterating on this. Links to the full piece and the dashboard are below.

Today, I'm releasing the first eval meant to test whether frontier models will help with authoritarian requests, or resist--the Dictatorship Eval. Headline finding: while some models resist direct authoritarian requests, they all comply with requests disguised as innocuous edits to codebases. As AI is woven into the government and so many parts of society, the biggest near-term risk for freedom isn't some scifi dictatorship of a runaway AI: it's people inside government or inside model companies using the technology to suppress or control us. Model companies understand this, and several of them (particularly Anthropic and OpenAI) have written explicit policies meant to prevent the models from going along with nefarious requests like these. But how well are these policies playing out in practice? Despite all the recent discussion of these issues around the conflict between Anthropic and the Pentagon, no one has systematically tested what the models actually do in these contexts, as opposed to what people in government and industry say they're supposed to do. That's what the Dictatorship Eval does. And the findings suggest we have a lot of work to do to align the policies with what really goes on in practice. It's hard to define what counts as an authoritarian request, so I'm open sourcing the whole library of scenarios I used so that others can improve on them. It's also hard to get an accurate picture of how the models might be used for authoritarian ends, because I can only test hypothetical requests using public-facing models, while the government and the model companies can obviously use internal models with different guardrails. But hopefully this work is a useful first step that gives us some sense of what's going on, and a sort of "lower bound" on how models comply with these requests. Finally: it's not obvious to me that the correct solution here is increasing the rate at which models refuse these requests. Do we really want models scanning our code and judging its moral value before agreeing to help us? Or should we double down on improving how we govern against authoritarianism at the societal level, while leaving the tools open to fulfilling most requests? The answer is probably in between. Just like we don't want the models to help create bioweapons, we probably do want them to explicitly refuse outrageous requests. But we probably also want to limit how often and how strongly they refuse and fall back on other means for guarding against their use for authoritarian ends. I'm super grateful to everyone who gave me feedback on this project along the way, especially Ethan BdM , Zhengdong , Connor Huff, and a bunch of folks at Anthropic. Looking forward to getting feedback from the community and iterating on this. Links to the full piece and the dashboard are below.

Andy Hall

33,905 次观看 • 4 个月前

Experiments in progress. The one on the right has been learning for ~3 hours, the one in the middle for ~1 hour, and the one on the left just started a few minutes ago. The initial motivation for making the physical Atari was just to commit ourselves to a subset of algorithms that can make progress in this setup. This commitment rules out algorithms that require billions of samples to learn (or worse, require multiple environments running in parallel). Atari games are simple enough that we should be able to show learning on them in a short amount of time with no prior knowledge. Since then, I've realized that this setup is also a good way to compare different paradigms in robotics in a principled way. These paradigms are sim2real, learning from tele-operated data, and learning directly on the robots. So far, I have observed that getting sim2real to work reliably is hard. It requires tweaks that don't scale. Policies that can play perfectly in simulation fall apart because of latencies and the messiness of the real world. These aspects could be modeled to improve the simulation, but not without sinking significant human engineering hours. I have higher hopes for learning from tele-operated data, but that requires a human to learn the task first. These experiments are on my to-do list. I have to learn to play some of the games well through the robot. I’m half-decent at playing Pong and Ms Pacman now. Learning directly on robots is looking like the most promising approach. This approach takes away pesky distribution shifts and makes it possible to have algorithms that continually improve with more data and time without any human intervention. It feels great to let experiments run overnight and wake up to find improved policies. With learning on robots, I should, in principle, be able to go on a long vacation and come back to find better policies for complex tasks beyond Atari games. Whether that is possible with current learning algorithms is a different question.

Experiments in progress. The one on the right has been learning for ~3 hours, the one in the middle for ~1 hour, and the one on the left just started a few minutes ago. The initial motivation for making the physical Atari was just to commit ourselves to a subset of algorithms that can make progress in this setup. This commitment rules out algorithms that require billions of samples to learn (or worse, require multiple environments running in parallel). Atari games are simple enough that we should be able to show learning on them in a short amount of time with no prior knowledge. Since then, I've realized that this setup is also a good way to compare different paradigms in robotics in a principled way. These paradigms are sim2real, learning from tele-operated data, and learning directly on the robots. So far, I have observed that getting sim2real to work reliably is hard. It requires tweaks that don't scale. Policies that can play perfectly in simulation fall apart because of latencies and the messiness of the real world. These aspects could be modeled to improve the simulation, but not without sinking significant human engineering hours. I have higher hopes for learning from tele-operated data, but that requires a human to learn the task first. These experiments are on my to-do list. I have to learn to play some of the games well through the robot. I’m half-decent at playing Pong and Ms Pacman now. Learning directly on robots is looking like the most promising approach. This approach takes away pesky distribution shifts and makes it possible to have algorithms that continually improve with more data and time without any human intervention. It feels great to let experiments run overnight and wake up to find improved policies. With learning on robots, I should, in principle, be able to go on a long vacation and come back to find better policies for complex tasks beyond Atari games. Whether that is possible with current learning algorithms is a different question.

Khurram Javed

52,110 次观看 • 8 个月前

Q: Esteban, new year, new rules, new car. You've had the first run out this morning at the Shakedown. What are the first impressions? How did the program go for you? Esteban Ocon: Yeah, feeling good. I think, frist of all, an unbelievable effort from the team really to put the car down, you know, at 9:20 this morning. But the was ready at 9:00 you know. We were waiting on a bit of a better track condition and a few things that we wanted to be perfect before we went out. But yeah, from Fiorano testing with Ollie to here there's been moving like people have climbed mountains really to make this car work and it's been really good. So, we are dealing with the plan, learning as it goes. Of course, it's a busy program that we have for the day. So you know, its gonna be difficult to complete it. But for the first real day of driving, I think so far it's going really well and we'll keep pushing to make sure that all the details are covered. But we have more days than normal which is a good thing. Q: Very early days, of course, but just how different are these cars to drive? Have you had a chance to play around a little bit with some of the new modes? Ocon: Yeah, it's very difficult, very complicated. I got lucky be able to do a lot of simulator days before we started the year, so we are pretty well set on that. Everything is clear, but yes, it's very complicated you know, for all of us. But I hope that this will be the same for everyone, because if it is we're in the same boat, so we'll see. Q: But I get the sense you're really relishing that challange and what about the priorities from here, the rest of the weel here at the shakedown? Ocon: Yeah, the aim is really to learn, to get mileage under the car, you know, see the weak points, what we have to improve really. First feel of things, so we are sure that we take the right development path and we are sure that we put the resources where it matters the most. Where it's the most bothering us, so you know, we'll try and put all that together for that end of the test. It's a long week, which is very good and then we have the chance to go back to Bahrain with hopefully, further step made, so that's the aim. #HaasF1 #F1Testing #F1

Q: Esteban, new year, new rules, new car. You've had the first run out this morning at the Shakedown. What are the first impressions? How did the program go for you? Esteban Ocon: Yeah, feeling good. I think, frist of all, an unbelievable effort from the team really to put the car down, you know, at 9:20 this morning. But the was ready at 9:00 you know. We were waiting on a bit of a better track condition and a few things that we wanted to be perfect before we went out. But yeah, from Fiorano testing with Ollie to here there's been moving like people have climbed mountains really to make this car work and it's been really good. So, we are dealing with the plan, learning as it goes. Of course, it's a busy program that we have for the day. So you know, its gonna be difficult to complete it. But for the first real day of driving, I think so far it's going really well and we'll keep pushing to make sure that all the details are covered. But we have more days than normal which is a good thing. Q: Very early days, of course, but just how different are these cars to drive? Have you had a chance to play around a little bit with some of the new modes? Ocon: Yeah, it's very difficult, very complicated. I got lucky be able to do a lot of simulator days before we started the year, so we are pretty well set on that. Everything is clear, but yes, it's very complicated you know, for all of us. But I hope that this will be the same for everyone, because if it is we're in the same boat, so we'll see. Q: But I get the sense you're really relishing that challange and what about the priorities from here, the rest of the weel here at the shakedown? Ocon: Yeah, the aim is really to learn, to get mileage under the car, you know, see the weak points, what we have to improve really. First feel of things, so we are sure that we take the right development path and we are sure that we put the resources where it matters the most. Where it's the most bothering us, so you know, we'll try and put all that together for that end of the test. It's a long week, which is very good and then we have the chance to go back to Bahrain with hopefully, further step made, so that's the aim. #HaasF1 #F1Testing #F1

Lucho Yoma

12,593 次观看 • 6 个月前

Welcome to the Lab of the Future! 🧬🤖 Excited to share LUMI-lab, out today in Cell — a self-driving platform that pairs an AI foundation model with a robotic lab to autonomously discover ionizable lipids (LNPs) for mRNA delivery. The core problem: Designing lipid nanoparticles (LNPs) is hard. The chemical space of ionizable lipids is vast, experimental cycles are slow, and — critically — historical LNP datasets are far too small to train a predictive model from scratch. Most AI approaches in this space hit a wall immediately: not enough data to learn from. Our solution: lab-in-the-loop foundation model learning. Instead of training on LNP data alone, LUMI starts as a transformer-based foundation model pretrained across broad chemical space, building rich molecular representations before it ever sees a single LNP experiment. Then it enters a closed loop with a robotic synthesis platform: predict → synthesize → assay → update. Each round of real wet-lab experiments fine-tunes the model, which then proposes smarter candidates for the next round. The lab isn't just validating AI predictions — it's actively teaching the model, continuously. What happened when we let it run: LUMI-lab autonomously synthesized and screened 1,700+ ionizable lipids in human bronchial epithelial cells. The top candidate — LUMI-6 — features a brominated lipid tail, a structural motif that had been largely overlooked in LNP design. LUMI found it without being told where to look. When formulated into LNPs and delivered intratracheally to mice, LUMI-6 achieved 20.3% gene editing efficiency in lung epithelial cells — a compelling result for one of the hardest-to-reach therapeutic targets, directly relevant to diseases like cystic fibrosis and alpha-1 antitrypsin deficiency. Why this matters beyond LNPs: This is a proof of concept for a broader thesis — that foundation model pretraining + active learning + robotic experimentation can overcome the data scarcity bottleneck that plagues AI-driven discovery in biology. You don't need a massive domain-specific dataset to start. You need a model that can generalize, a lab that can generate the right data, and a loop that connects them. Huge congratulations to first authors Yue Xu, Haotian Cui, and Kuan Pang, and to the entire Bowen LI team. Grateful to our collaborators at University Health Network and Leslie Dan Faculty of Pharmacy, and to Princess Margaret Cancer Centre Research Princess Margaret Cancer Centre Research. 📄 Paper:

Welcome to the Lab of the Future! 🧬🤖 Excited to share LUMI-lab, out today in Cell — a self-driving platform that pairs an AI foundation model with a robotic lab to autonomously discover ionizable lipids (LNPs) for mRNA delivery. The core problem: Designing lipid nanoparticles (LNPs) is hard. The chemical space of ionizable lipids is vast, experimental cycles are slow, and — critically — historical LNP datasets are far too small to train a predictive model from scratch. Most AI approaches in this space hit a wall immediately: not enough data to learn from. Our solution: lab-in-the-loop foundation model learning. Instead of training on LNP data alone, LUMI starts as a transformer-based foundation model pretrained across broad chemical space, building rich molecular representations before it ever sees a single LNP experiment. Then it enters a closed loop with a robotic synthesis platform: predict → synthesize → assay → update. Each round of real wet-lab experiments fine-tunes the model, which then proposes smarter candidates for the next round. The lab isn't just validating AI predictions — it's actively teaching the model, continuously. What happened when we let it run: LUMI-lab autonomously synthesized and screened 1,700+ ionizable lipids in human bronchial epithelial cells. The top candidate — LUMI-6 — features a brominated lipid tail, a structural motif that had been largely overlooked in LNP design. LUMI found it without being told where to look. When formulated into LNPs and delivered intratracheally to mice, LUMI-6 achieved 20.3% gene editing efficiency in lung epithelial cells — a compelling result for one of the hardest-to-reach therapeutic targets, directly relevant to diseases like cystic fibrosis and alpha-1 antitrypsin deficiency. Why this matters beyond LNPs: This is a proof of concept for a broader thesis — that foundation model pretraining + active learning + robotic experimentation can overcome the data scarcity bottleneck that plagues AI-driven discovery in biology. You don't need a massive domain-specific dataset to start. You need a model that can generalize, a lab that can generate the right data, and a loop that connects them. Huge congratulations to first authors Yue Xu, Haotian Cui, and Kuan Pang, and to the entire Bowen LI team. Grateful to our collaborators at University Health Network and Leslie Dan Faculty of Pharmacy, and to Princess Margaret Cancer Centre Research Princess Margaret Cancer Centre Research. 📄 Paper:

Bo Wang

57,510 次观看 • 5 个月前

The $AEGIS DApp portal is now open to all: 🛡️ At Aegis, we believe in empowering the blockchain full of security, transparency and innovation. The Aegis Dapp has been under development for several months prior to the launch of $AEGIS and with that we have been able to build what we believe has the potential to change how users go about their day to day security. We are thrilled to share our progress and truly exciting news with you all. 🎯 First things first, at Aegis, we want to make it clear that the value of what we seek to bring to security across the blockchain, comes from our big vision, our strong team, and our commitment to long-term goals. ℹ️ Let’s kick this off with some information that is constantly happening, which is behind the scenes. Our full team is dedicated to the opportunity that lays ahead of us with becoming the leading voice/name for security, grasping every aspect with innovation, hard work, passion and commitment to see this sector grow. Everyone is aware of how important security is, a heartwarming mention to Messari for including us on how they see this sector growing rapidly and pushing a 10 Billion evaluation. We take that recognition with full responsibility and gratitude as we've been working hard on some really powerful stuff that could change the game for our industry. If you read the title and report itself, I’m sure that’ll give you some insight to what’s coming, and to the vast extent of what you can expect Aegis to be working towards. —> 🤝 This comes from teaming up with others within this sector and coming up with new tech to projects driven by our community, within the pipeline you can be confident that what we are building will push the cryptocurrency industry as a whole into a better future, the magnitude to what Aegis brings will not stop until we can confidently say, “Negative security reports across the blockchain are at an all time low, thousands of users are satisfied that Aegis is protecting them and their assets.” We're sticking to our vision no matter what the market does or whatever else comes our way. We plan to build what we set out to and we will see to it that our ecosystem is met. We've been working on some pretty amazing products that will be available within our Dapp, let’s go over what we offer: * AI Audits * Live Monitoring * Penetration Testing * Bug Bounties * Live Watchdog * Token analytics for everyday users, developers, teams, auditors, institutions, investors. ⬇️ Let’s break it down for you in some simple steps: AI AUDITS: We have trained our LLM models as AI AGENTS, these consist of 3 people ( AI AGENTS ) for the audits that are performed. - Audit - Reviewer - Judge Each one analyzes with a different personality, let’s check what personalities our AI AGENTS consist of: 3 different perspective auditors. 1 - Fine-tuned model x amount reads the code and generates the audit. ✅ 2 - Model x amount reviews the code and fact checks thoroughly. ✅ 3 - Model x amount ranks the code based on the severity outcome. ✅ ⌚️ Live Monitoring/Watchdog: The Live Monitoring/Watchdog system is designed to provide real-time surveillance of smart contracts, ensuring the detection and prevention of any potentially harmful transactions or malicious activities. Through the utilization of an AI Agent model, the system is trained to proactively identify and thwart suspicious behavior, thereby safeguarding the integrity of the smart contracts. Also, a paid sophisticated threat detection model is available for more intricate protocols and Dapps, offering an advanced level of protection against potential threats. This proactive approach is crucial in mitigating the risk of exploitation and ensuring the security of the smart contract ecosystem. 🖊️ Pen Testing: Our platform offers Pen Testing services to developers, providing a controlled environment for whitehat hackers to simulate attacks and identify vulnerabilities in smart contracts and protocols. In addition to human whitehat hackers, our AI Agents function as Red and Blue teams, actively engaging in simulated attacks to stress-test protocols and identify potential weaknesses. This comprehensive approach allows developers to proactively identify and address security issues, ultimately enhancing the robustness and resilience of their projects. 🕷️ Bug Bounties: Our Bug Bounty listing platform provides developers with the opportunity to list their protocols and offer bounties to white hat hackers for identifying vulnerabilities. By aggregating millions of bounties from various platforms and utilizing AI tools, we streamline the testing process, reducing up to 80% of the workload typically associated with security testing. This allows developers to efficiently identify and address potential vulnerabilities in their protocols, ultimately enhancing the overall security and resilience of their projects. 🪙 And lot more token analytics features for regular users, this will give you the opportunity to explore our Dapp for yourself and have some fun diving into the security platform of the future! I’m sure you’re excited to try it all out yourself, which is why we have some exciting news to bring to the #Guardians of the blockchain! But just before you continue the read and see the beans have been spilled, we have to take this opportunity to share with you that this large step to becoming a security leader is but only 20% of what we have revealed. This will be at the core of what Aegis stands for and hopes to achieve. The focus here is upon our Dapp, and in time we will slowly bring forward information/updates regarding segments of what makes Aegis a force to be reckoned with. Now that you’re fired up and excited to all of the announcements to come, let’s get to the news you’ve been waiting for! 🎉 We’re spilling the good news, and are happy to say we are now set for public release! The team at Aegis are overwhelmed with the development, support from teams, community, partners and more on what we believe to be an institutional-grade product. But the fun doesn’t stop there, this marks the start of what we aim to become, as it will take time and cycles to become better and better. Constant advancements will be set in place to attain the goal of achieving blockchain security. A statement from our CEO- Brian Hunt: “I can confirm from the security conferences I attended with Centralized security firms Peckshield, Hacken, Certik, BlockSec presentations, they are trying to achieve something similar and it will take them years. Decentralized AI for Security!” This initial drop of our dapp will be to get users signed up to gain access, in which we’ll whitelist users to get the ball rolling. 📣 To end this segment, let’s get the party started with the long awaited Aegis Ai Security Dapp and sign up now!

The $AEGIS DApp portal is now open to all: 🛡️ At Aegis, we believe in empowering the blockchain full of security, transparency and innovation. The Aegis Dapp has been under development for several months prior to the launch of $AEGIS and with that we have been able to build what we believe has the potential to change how users go about their day to day security. We are thrilled to share our progress and truly exciting news with you all. 🎯 First things first, at Aegis, we want to make it clear that the value of what we seek to bring to security across the blockchain, comes from our big vision, our strong team, and our commitment to long-term goals. ℹ️ Let’s kick this off with some information that is constantly happening, which is behind the scenes. Our full team is dedicated to the opportunity that lays ahead of us with becoming the leading voice/name for security, grasping every aspect with innovation, hard work, passion and commitment to see this sector grow. Everyone is aware of how important security is, a heartwarming mention to Messari for including us on how they see this sector growing rapidly and pushing a 10 Billion evaluation. We take that recognition with full responsibility and gratitude as we've been working hard on some really powerful stuff that could change the game for our industry. If you read the title and report itself, I’m sure that’ll give you some insight to what’s coming, and to the vast extent of what you can expect Aegis to be working towards. —> 🤝 This comes from teaming up with others within this sector and coming up with new tech to projects driven by our community, within the pipeline you can be confident that what we are building will push the cryptocurrency industry as a whole into a better future, the magnitude to what Aegis brings will not stop until we can confidently say, “Negative security reports across the blockchain are at an all time low, thousands of users are satisfied that Aegis is protecting them and their assets.” We're sticking to our vision no matter what the market does or whatever else comes our way. We plan to build what we set out to and we will see to it that our ecosystem is met. We've been working on some pretty amazing products that will be available within our Dapp, let’s go over what we offer: * AI Audits * Live Monitoring * Penetration Testing * Bug Bounties * Live Watchdog * Token analytics for everyday users, developers, teams, auditors, institutions, investors. ⬇️ Let’s break it down for you in some simple steps: AI AUDITS: We have trained our LLM models as AI AGENTS, these consist of 3 people ( AI AGENTS ) for the audits that are performed. - Audit - Reviewer - Judge Each one analyzes with a different personality, let’s check what personalities our AI AGENTS consist of: 3 different perspective auditors. 1 - Fine-tuned model x amount reads the code and generates the audit. ✅ 2 - Model x amount reviews the code and fact checks thoroughly. ✅ 3 - Model x amount ranks the code based on the severity outcome. ✅ ⌚️ Live Monitoring/Watchdog: The Live Monitoring/Watchdog system is designed to provide real-time surveillance of smart contracts, ensuring the detection and prevention of any potentially harmful transactions or malicious activities. Through the utilization of an AI Agent model, the system is trained to proactively identify and thwart suspicious behavior, thereby safeguarding the integrity of the smart contracts. Also, a paid sophisticated threat detection model is available for more intricate protocols and Dapps, offering an advanced level of protection against potential threats. This proactive approach is crucial in mitigating the risk of exploitation and ensuring the security of the smart contract ecosystem. 🖊️ Pen Testing: Our platform offers Pen Testing services to developers, providing a controlled environment for whitehat hackers to simulate attacks and identify vulnerabilities in smart contracts and protocols. In addition to human whitehat hackers, our AI Agents function as Red and Blue teams, actively engaging in simulated attacks to stress-test protocols and identify potential weaknesses. This comprehensive approach allows developers to proactively identify and address security issues, ultimately enhancing the robustness and resilience of their projects. 🕷️ Bug Bounties: Our Bug Bounty listing platform provides developers with the opportunity to list their protocols and offer bounties to white hat hackers for identifying vulnerabilities. By aggregating millions of bounties from various platforms and utilizing AI tools, we streamline the testing process, reducing up to 80% of the workload typically associated with security testing. This allows developers to efficiently identify and address potential vulnerabilities in their protocols, ultimately enhancing the overall security and resilience of their projects. 🪙 And lot more token analytics features for regular users, this will give you the opportunity to explore our Dapp for yourself and have some fun diving into the security platform of the future! I’m sure you’re excited to try it all out yourself, which is why we have some exciting news to bring to the #Guardians of the blockchain! But just before you continue the read and see the beans have been spilled, we have to take this opportunity to share with you that this large step to becoming a security leader is but only 20% of what we have revealed. This will be at the core of what Aegis stands for and hopes to achieve. The focus here is upon our Dapp, and in time we will slowly bring forward information/updates regarding segments of what makes Aegis a force to be reckoned with. Now that you’re fired up and excited to all of the announcements to come, let’s get to the news you’ve been waiting for! 🎉 We’re spilling the good news, and are happy to say we are now set for public release! The team at Aegis are overwhelmed with the development, support from teams, community, partners and more on what we believe to be an institutional-grade product. But the fun doesn’t stop there, this marks the start of what we aim to become, as it will take time and cycles to become better and better. Constant advancements will be set in place to attain the goal of achieving blockchain security. A statement from our CEO- Brian Hunt: “I can confirm from the security conferences I attended with Centralized security firms Peckshield, Hacken, Certik, BlockSec presentations, they are trying to achieve something similar and it will take them years. Decentralized AI for Security!” This initial drop of our dapp will be to get users signed up to gain access, in which we’ll whitelist users to get the ball rolling. 📣 To end this segment, let’s get the party started with the long awaited Aegis Ai Security Dapp and sign up now!

AEGIS AI

128,006 次观看 • 2 年前

🎉 The best way to start the week is to find out that our MedSAM is finally published today in Nature Communications! **Segment anything in medical images** Paper: arXiv: Data & Code: MedSAM is the first promotable foundation model for medical image segmentation. **Highlights**: ⭐ Before its formal publication, we have received 220 citations and 1400+ GitHub stars 🙏🙏❤️‍🔥❤️‍🔥❤️‍🔥 📊 We curated a large-scale medical image dataset with 1,570,263 image-mask pairs, covering 10 imaging modalities and over 30 cancer types. 🚀 Built on top of SAM (AI at Meta ) with transfer learning, we have significantly enhanced its segmentation performance of medical images. 📈 Comprehensive evaluations of 86 internal validation tasks and 60 external validation tasks demonstrate its better accuracy and robustness than modality-wise specialist models. **What is Next? --- Clinical Translation!!** 🍕Our next goal is to make the model deployable on laptops (CPUs) or other edge devices without reliance on GPUs. We have distilled a lightweight model, LiteMedSAM, offering a speed boost of 10x while maintaining accuracy. Plus, we have integrated it into the 3D Slicer plugin, providing an efficient tool for medical image segmentation. 🌐 To further promote developments in this field, we organize a competition on #CVPR2026: Segment Anything in Medical Images on Laptop! An out-of-the-box baseline has been released to reduce the entry barriers. Welcome to join us to push the boundary further: 🙏 Massive thanks to MetaAI AI at Meta for their open-source project SAM and many reviewers/users for their invaluable feedback. A huge shoutout to my postdoc Jun Ma (JunMa) for his leadership on this project!! UHN AI Hub Vector Institute Peter Munk Cardiac Centre AI Department of Laboratory Medicine & Pathobiology U of T Department of Computer Science University of Toronto University Health Network Brad Wouters 🇨🇦 Barry Rubin MD, PhD, FRCSC Shaf Keshavjee

🎉 The best way to start the week is to find out that our MedSAM is finally published today in Nature Communications! Segment anything in medical images Paper: arXiv: Data & Code: MedSAM is the first promotable foundation model for medical image segmentation. Highlights: ⭐ Before its formal publication, we have received 220 citations and 1400+ GitHub stars 🙏🙏❤️‍🔥❤️‍🔥❤️‍🔥 📊 We curated a large-scale medical image dataset with 1,570,263 image-mask pairs, covering 10 imaging modalities and over 30 cancer types. 🚀 Built on top of SAM (AI at Meta ) with transfer learning, we have significantly enhanced its segmentation performance of medical images. 📈 Comprehensive evaluations of 86 internal validation tasks and 60 external validation tasks demonstrate its better accuracy and robustness than modality-wise specialist models. What is Next? --- Clinical Translation!! 🍕Our next goal is to make the model deployable on laptops (CPUs) or other edge devices without reliance on GPUs. We have distilled a lightweight model, LiteMedSAM, offering a speed boost of 10x while maintaining accuracy. Plus, we have integrated it into the 3D Slicer plugin, providing an efficient tool for medical image segmentation. 🌐 To further promote developments in this field, we organize a competition on #CVPR2026: Segment Anything in Medical Images on Laptop! An out-of-the-box baseline has been released to reduce the entry barriers. Welcome to join us to push the boundary further: 🙏 Massive thanks to MetaAI AI at Meta for their open-source project SAM and many reviewers/users for their invaluable feedback. A huge shoutout to my postdoc Jun Ma (JunMa) for his leadership on this project!! UHN AI Hub Vector Institute Peter Munk Cardiac Centre AI Department of Laboratory Medicine & Pathobiology U of T Department of Computer Science University of Toronto University Health Network Brad Wouters 🇨🇦 Barry Rubin MD, PhD, FRCSC Shaf Keshavjee

Bo Wang

140,229 次观看 • 2 年前

In 1968, George Land (with Beth Jarman) conducted a research study to test the creativity of 1,600 children ranging in ages from three-to-five years old. This was the same creativity test he devised for NASA to help select innovative engineers and scientists. The assessment worked so well he decided to try it on children. He then re-tested the same children at 10 years of age, and again at 15 years of age (a longitudinal study). The test was designed to indicate how well someone could look at a issue and devise new, different, innovative ways to address it. To do that, they asked children at these various ages to come up with ways to use a paperclip. The results may surprise you. •Creativity scores amongst 5-year old’s: 98% •Creativity scores amongst 10-year old’s: 30% •Creativity scores amongst 15-year old’s: 12% •Given to 280,000 adults (average age of 31): 2% The results are not as we would expect, are they? The proportion of people who scored at the “Genius Level” decreased with age. We might think that it should increase with our level of education. But clearly not. Why is this? Well, it seems we have employed a system that educates the genius right out of people. And how have we done this? The industrial revolution began in Great Britain in the mid 1700's and by 1760 in the U.S was gaining steam - literally. Steam power was the catalyst. This first industrial revolution exploded from about 1760 to 1840. It was followed by the age of science and mass production, and then the digital revolution. We are now at the beginning of the next phase of dramatic technological expansion and social change—the Fourth Industrial Revolution. Coinciding with the first industrial revolution was an educational revolution, Public Education. The system was created in the late 1600's and in the 1700's was developed into the system we still use today. It was designed to meet the challenges of the first industrial revolution. But, when that revolution gave way to the second, and third, and fourth, our educational system did not. It was devoted to the original system. Alongside the first industrial revolution, it has continued to insist on manufacturing the same student, over and over, tweaking content for new technologies. So, today's education system produces student conformity just as any good industrial manufacturing process would do with its product. But not student creativity, diversity or individual expression, which fits the needs of our day. As a result, ideas for paperclips decline and genius dries up. This is not the path to a world where AI is built into everything. It is vital to bring our children back to the path of 98% genius level as presented by Dr. Land. We can also bring back us adults back from the institutional learning of schools and universities. The reality is we simply have no choice as there is no path forward if human individuality and creativity is not asserted as our primary function. This is and always has been the path of humanity. We just need to remember to remember.

In 1968, George Land (with Beth Jarman) conducted a research study to test the creativity of 1,600 children ranging in ages from three-to-five years old. This was the same creativity test he devised for NASA to help select innovative engineers and scientists. The assessment worked so well he decided to try it on children. He then re-tested the same children at 10 years of age, and again at 15 years of age (a longitudinal study). The test was designed to indicate how well someone could look at a issue and devise new, different, innovative ways to address it. To do that, they asked children at these various ages to come up with ways to use a paperclip. The results may surprise you. •Creativity scores amongst 5-year old’s: 98% •Creativity scores amongst 10-year old’s: 30% •Creativity scores amongst 15-year old’s: 12% •Given to 280,000 adults (average age of 31): 2% The results are not as we would expect, are they? The proportion of people who scored at the “Genius Level” decreased with age. We might think that it should increase with our level of education. But clearly not. Why is this? Well, it seems we have employed a system that educates the genius right out of people. And how have we done this? The industrial revolution began in Great Britain in the mid 1700's and by 1760 in the U.S was gaining steam - literally. Steam power was the catalyst. This first industrial revolution exploded from about 1760 to 1840. It was followed by the age of science and mass production, and then the digital revolution. We are now at the beginning of the next phase of dramatic technological expansion and social change—the Fourth Industrial Revolution. Coinciding with the first industrial revolution was an educational revolution, Public Education. The system was created in the late 1600's and in the 1700's was developed into the system we still use today. It was designed to meet the challenges of the first industrial revolution. But, when that revolution gave way to the second, and third, and fourth, our educational system did not. It was devoted to the original system. Alongside the first industrial revolution, it has continued to insist on manufacturing the same student, over and over, tweaking content for new technologies. So, today's education system produces student conformity just as any good industrial manufacturing process would do with its product. But not student creativity, diversity or individual expression, which fits the needs of our day. As a result, ideas for paperclips decline and genius dries up. This is not the path to a world where AI is built into everything. It is vital to bring our children back to the path of 98% genius level as presented by Dr. Land. We can also bring back us adults back from the institutional learning of schools and universities. The reality is we simply have no choice as there is no path forward if human individuality and creativity is not asserted as our primary function. This is and always has been the path of humanity. We just need to remember to remember.

Brian Roemmele

153,528 次观看 • 2 年前

$Introducing LifeGPT, showing that LLMs can simulate complex, Turing-complete systems like Conway's Game of Life with near-perfect accuracy—no prior topology needed.🌐This unlocks new potential for AI in modeling self-organizing systems in biology, materials science, & beyond.🔬🤖 #AI #LifeGPT. Cellular Automata (CA), like Conway's Game of Life ("Life"), are computationally irreducible, meaning their evolution is difficult to predict without an a-priori understanding of the rules of the game, including the topology on which it is played. LifeGPT is a topology-agnostic generative model that learns the rules of Life without prior knowledge of its grid structure or boundary conditions, from only a tiny number of game states. The success in simulating Life suggests promising avenues for scientific discovery, particularly in bridging the gap between AI, artificial life, and real-world biological systems, for both forward and inverse problems. The potential for universal computation within generative AI, including LLMs, through approaches like LifeGPT, represents an exciting area for future research, especially when combined with reinforcement learning. Model Convergence: LifeGPT exhibits rapid convergence during training, achieving high accuracy in predicting next-game-states. We attribute the non-zero cross-entropy loss to the lack of causal relationships within randomly generated ICs. Accuracy & Temperature: LifeGPT achieves near-perfect accuracy, particularly at lower sampling temperatures, but can be continually tuned towards higher creativity to discover patterns that the original ruleset would not be able to produce. This finding highlights the trade-off between model creativity (higher temperature) and accuracy in deterministic predictions, with high relevance to model real-world dynamical systems for which no closed-form rulesets exist. Zero/Few-Shot Learning: Trained on a small fraction of possible initial conditions, LifeGPT demonstrates strong zero/few-shot learning, accurately simulating Life for unseen initial conditions. However, rare prediction errors highlight that LifeGPT approximates rather than perfectly replicates the Life algorithm. Autoregressive Autoregressor: A recursive implementation of LifeGPT demonstrates the model's ability to simulate Life over multiple timesteps. LifeGPT is topology-agnostic with respect to its training data and our results show that a GPT model is capable of capturing the deterministic rules of a Turing-complete system with near-perfect accuracy, given sufficiently diverse training data. The work showcases the possibility for future models to synthesize stochastic generative capabilities with deterministic computational capabilities. Link to code, paper, etc. below. Podcast generated using #NotebookLM. LAMM@MIT DMSE at MIT$

Introducing LifeGPT, showing that LLMs can simulate complex, Turing-complete systems like Conway's Game of Life with near-perfect accuracy—no prior topology needed.🌐This unlocks new potential for AI in modeling self-organizing systems in biology, materials science, & beyond.🔬🤖 #AI #LifeGPT. Cellular Automata (CA), like Conway's Game of Life ("Life"), are computationally irreducible, meaning their evolution is difficult to predict without an a-priori understanding of the rules of the game, including the topology on which it is played. LifeGPT is a topology-agnostic generative model that learns the rules of Life without prior knowledge of its grid structure or boundary conditions, from only a tiny number of game states. The success in simulating Life suggests promising avenues for scientific discovery, particularly in bridging the gap between AI, artificial life, and real-world biological systems, for both forward and inverse problems. The potential for universal computation within generative AI, including LLMs, through approaches like LifeGPT, represents an exciting area for future research, especially when combined with reinforcement learning. Model Convergence: LifeGPT exhibits rapid convergence during training, achieving high accuracy in predicting next-game-states. We attribute the non-zero cross-entropy loss to the lack of causal relationships within randomly generated ICs. Accuracy & Temperature: LifeGPT achieves near-perfect accuracy, particularly at lower sampling temperatures, but can be continually tuned towards higher creativity to discover patterns that the original ruleset would not be able to produce. This finding highlights the trade-off between model creativity (higher temperature) and accuracy in deterministic predictions, with high relevance to model real-world dynamical systems for which no closed-form rulesets exist. Zero/Few-Shot Learning: Trained on a small fraction of possible initial conditions, LifeGPT demonstrates strong zero/few-shot learning, accurately simulating Life for unseen initial conditions. However, rare prediction errors highlight that LifeGPT approximates rather than perfectly replicates the Life algorithm. Autoregressive Autoregressor: A recursive implementation of LifeGPT demonstrates the model's ability to simulate Life over multiple timesteps. LifeGPT is topology-agnostic with respect to its training data and our results show that a GPT model is capable of capturing the deterministic rules of a Turing-complete system with near-perfect accuracy, given sufficiently diverse training data. The work showcases the possibility for future models to synthesize stochastic generative capabilities with deterministic computational capabilities. Link to code, paper, etc. below. Podcast generated using #NotebookLM. LAMM@MIT DMSE at MIT

Markus J. Buehler

114,217 次观看 • 1 年前