Loading video...

Video Failed to Load

Go Home

Russian scientists breakthrough in safer rocket design with model that predicts when space fuel ignites The new mathematical model calculates ignition delay time with high accuracy (within 10–15% error) for hypergolic fuels — propellants that ignite on contact without a spark or heat 🔸The model accounts for heat exchange,...

SputnikInt's profile picture

Sputnik

500,053 subscribers

23,699 views • 7 days ago •via X (Twitter)

Science & TechnologyHealth & WellnessNews & Politics
LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Building on the previous paper, in this study we compare a continuous “smooth return” S2>S1 model with an event-driven one, where long periods of relative calm are punctuated by short, intense episodes of global reorganisation. Both models cover the same time window. Neither uses archaeological data in its construction. When compared against where early humans and early civilizations actually appear and persist, the difference is statistically robust. The smooth model behaves like background noise. The event-driven model lines up in time and space far better than chance allows, even after aggressive temporal and spatial randomization tests. Statistically, the event-driven model lines up with where and when early civilizations appear far better than a smooth, continuous model, even after we randomize both timing and location to test what could arise by chance. The event timeline itself was built independently from well-known late-glacial disruptions - such as Heinrich events, meltwater pulses, and abrupt deglacial transitions - rather than from any archaeological data. Nothing here claims that specific events caused specific cultures. It does suggest that history may not unfold on a smooth clock. Human societies seem to flourish during recovery phases between disruptions, not during the disruptions themselves. The animation contrasts the two return models. Draft paper : Source & Results : (coming soon)
0:16

Building on the previous paper, in this study we compare a continuous “smooth return” S2>S1 model with an event-driven one, where long periods of relative calm are punctuated by short, intense episodes of global reorganisation. Both models cover the same time window. Neither uses archaeological data in its construction. When compared against where early humans and early civilizations actually appear and persist, the difference is statistically robust. The smooth model behaves like background noise. The event-driven model lines up in time and space far better than chance allows, even after aggressive temporal and spatial randomization tests. Statistically, the event-driven model lines up with where and when early civilizations appear far better than a smooth, continuous model, even after we randomize both timing and location to test what could arise by chance. The event timeline itself was built independently from well-known late-glacial disruptions - such as Heinrich events, meltwater pulses, and abrupt deglacial transitions - rather than from any archaeological data. Nothing here claims that specific events caused specific cultures. It does suggest that history may not unfold on a smooth clock. Human societies seem to flourish during recovery phases between disruptions, not during the disruptions themselves. The animation contrasts the two return models. Draft paper : Source & Results : (coming soon)

Craig Stone

10,899 views • 6 months ago

So remember that time machine I told you about? I managed to sneek out another video from my visit to the future.. this time I captured video of a 3d artist creating assets for a short commercial. They sculpted the bird in the video with nothing but two controllers which they used to manipulate generative matter in 3d. The AI model changed in realtime and adopted their style of work, both based on the work they did on the scene, and on reference images they gave the system. Seeing the 3d model transform in realtime as they were working on it was incredible. When it started moving and reacted to their instructions, I knew it was time to return to the present :D #art #ai
0:16

So remember that time machine I told you about? I managed to sneek out another video from my visit to the future.. this time I captured video of a 3d artist creating assets for a short commercial. They sculpted the bird in the video with nothing but two controllers which they used to manipulate generative matter in 3d. The AI model changed in realtime and adopted their style of work, both based on the work they did on the scene, and on reference images they gave the system. Seeing the 3d model transform in realtime as they were working on it was incredible. When it started moving and reacted to their instructions, I knew it was time to return to the present :D #art #ai

Martin Nebelong

22,044 views • 2 years ago

$IKA 🪂 Airdrop – 60-second brief - covering my first $SUI ecco drop 👇 ➡️Allocation 🔸 600 M $IKA (6 % of 10 B) allocated for a “community drop” at mainnet - May 8 tokenomics post. ➡️ What’s Ika Ika「🦑」 )? 🔸Sub-second parallel MPC network (~10 k sigs/s) that lets Sui dApps sign BTC / ETH / SOL txs natively - no bridges, no custodians. How to farm (core tasks visible in Ink Sack now) 1⃣Stake one MF Squid NFT → ~100 Droplets / day. 2⃣Lock iSUI (10 iSUI = 5 Droplets / day; 90 % withdrawable any time). 3⃣Use Squid Keys (earned by leveling the NFT) to unlock bonus Droplets - each key ≥ 1 Droplet, jackpot up to 10 M. 4⃣Pre-generate a dWallet (one-click, ~0.13 SUI gas) for an instant 50 Droplets. Tasks 5-8 = sign a msg + submit Sui, BTC & EVM addrs for extra crumbs. Everything lives at the official task list mirrors the above.  ➡️Why care? 🔸Allocation is public & fixed. 🔸Testnet is live; several Sui builders already integrating. 🔸Backers include Sui Foundation, DCG, FalconX — > $21 M in runway. ➡️Unknowns / risks 🔸Snapshot & TGE dates still unannounced 🔸Droplet-to-token curve & vesting schedule remain redacted. 🔸As hype grows, watch for fake “Ink Sack” URLs; the only legit one ends .wal.app. 🔸You have to use $SUI or real funds. NFA. DYOR. I'm not asking you, telling you to, or suggesting anything. Not Financial Advice. For entertainment purposes only. ➡️Why people spam GiveRep under Ika threads? GiveRep is a SocialFi bot on Sui; tagging it earns REP points for a separate airdrop. Farmers double-dip: Droplets for $IKA and REP for the future $REP drop. 🔸🔸Stake $SUI with our Sponsor Cosmos | Everstake and support Airdrops. 🔸🔸 Sources: 🔗Links ➡️Ink Sack task board: ➡️Swap SUI ↔ iSUI: ➡️MF Squid NFT marketplace: DYOR! NFA! Full content policy: 🫡
0:13

$IKA 🪂 Airdrop – 60-second brief - covering my first $SUI ecco drop 👇 ➡️Allocation 🔸 600 M $IKA (6 % of 10 B) allocated for a “community drop” at mainnet - May 8 tokenomics post. ➡️ What’s Ika Ika「🦑」 )? 🔸Sub-second parallel MPC network (~10 k sigs/s) that lets Sui dApps sign BTC / ETH / SOL txs natively - no bridges, no custodians. How to farm (core tasks visible in Ink Sack now) 1⃣Stake one MF Squid NFT → ~100 Droplets / day. 2⃣Lock iSUI (10 iSUI = 5 Droplets / day; 90 % withdrawable any time). 3⃣Use Squid Keys (earned by leveling the NFT) to unlock bonus Droplets - each key ≥ 1 Droplet, jackpot up to 10 M. 4⃣Pre-generate a dWallet (one-click, ~0.13 SUI gas) for an instant 50 Droplets. Tasks 5-8 = sign a msg + submit Sui, BTC & EVM addrs for extra crumbs. Everything lives at the official task list mirrors the above.  ➡️Why care? 🔸Allocation is public & fixed. 🔸Testnet is live; several Sui builders already integrating. 🔸Backers include Sui Foundation, DCG, FalconX — > $21 M in runway. ➡️Unknowns / risks 🔸Snapshot & TGE dates still unannounced 🔸Droplet-to-token curve & vesting schedule remain redacted. 🔸As hype grows, watch for fake “Ink Sack” URLs; the only legit one ends .wal.app. 🔸You have to use $SUI or real funds. NFA. DYOR. I'm not asking you, telling you to, or suggesting anything. Not Financial Advice. For entertainment purposes only. ➡️Why people spam GiveRep under Ika threads? GiveRep is a SocialFi bot on Sui; tagging it earns REP points for a separate airdrop. Farmers double-dip: Droplets for $IKA and REP for the future $REP drop. 🔸🔸Stake $SUI with our Sponsor Cosmos | Everstake and support Airdrops. 🔸🔸 Sources: 🔗Links ➡️Ink Sack task board: ➡️Swap SUI ↔ iSUI: ➡️MF Squid NFT marketplace: DYOR! NFA! Full content policy: 🫡

Airdrops

17,331 views • 1 year ago

🚨🇷🇺 WEST SHOCKED: RUSSIA JUST BUILT ITS FIRST CHIP MACHINE The West tried to cripple Russia with sanctions and cut off advanced chips. But Moscow just put its very first domestic photolithography machine — the Progress STP-350 — on open sale for about 400 million rubles. 🔸 This machine makes 350nm chips — bigger, tougher transistors that are perfect for military use. They resist radiation, Electromagnetic Pulse (EMP) attacks, extreme heat/cold, vibration, and high voltage (up to 100V) where super-small modern chips fail. 🔸 Perfect for triple-redundant military circuits (three copies of the same chip working together) that never fail even if one gets hit by cosmic rays or EMP pulse. 🔸 Handles extreme battlefield conditions: huge temperature swings, constant vibration, high voltage up to 100 Volts — things impossible on modern super-thin processes. 🔸 Uses a modern solid-state laser (365 nm) instead of old mercury lamps. It can process up to 63 silicon wafers per hour (150-200 mm size) and lasts much longer — up to 10,000 hours. 🔸 Developed since 2021 with help from Belarus company Planar — cutting Russia’s tech gap from 40-50 years down to about 30 years. 🔸 Ideal for critical defense systems: control units, engines, power supplies in missiles, planes, and radars that need reliability first, not maximum speed. 🔸 Foreign versions cost 2-3 times more. Tiny modern nodes are perfect for phones, but terrible for military use. 350nm is a mature, battle-proven tech that delivers superior reliability, radiation resistance, high voltage tolerance, and durability — exactly what defense systems and civilian sectors (cars, medicine, comms) actually need. Did Western sanctions actually make Russia stronger?
0:18

🚨🇷🇺 WEST SHOCKED: RUSSIA JUST BUILT ITS FIRST CHIP MACHINE The West tried to cripple Russia with sanctions and cut off advanced chips. But Moscow just put its very first domestic photolithography machine — the Progress STP-350 — on open sale for about 400 million rubles. 🔸 This machine makes 350nm chips — bigger, tougher transistors that are perfect for military use. They resist radiation, Electromagnetic Pulse (EMP) attacks, extreme heat/cold, vibration, and high voltage (up to 100V) where super-small modern chips fail. 🔸 Perfect for triple-redundant military circuits (three copies of the same chip working together) that never fail even if one gets hit by cosmic rays or EMP pulse. 🔸 Handles extreme battlefield conditions: huge temperature swings, constant vibration, high voltage up to 100 Volts — things impossible on modern super-thin processes. 🔸 Uses a modern solid-state laser (365 nm) instead of old mercury lamps. It can process up to 63 silicon wafers per hour (150-200 mm size) and lasts much longer — up to 10,000 hours. 🔸 Developed since 2021 with help from Belarus company Planar — cutting Russia’s tech gap from 40-50 years down to about 30 years. 🔸 Ideal for critical defense systems: control units, engines, power supplies in missiles, planes, and radars that need reliability first, not maximum speed. 🔸 Foreign versions cost 2-3 times more. Tiny modern nodes are perfect for phones, but terrible for military use. 350nm is a mature, battle-proven tech that delivers superior reliability, radiation resistance, high voltage tolerance, and durability — exactly what defense systems and civilian sectors (cars, medicine, comms) actually need. Did Western sanctions actually make Russia stronger?

NewRulesGeopolitics

67,272 views • 2 months ago

Big moment for text-to-speech. Qwen just open-sourced a text-to-speech model that lets you clone voices, design new ones, and control speech using natural language. Let me explain what I mean: You can literally tell it "speak in a cheerful tone with slight nervousness," and it actually does that. No complex audio engineering needed. What makes this special: - 3-second voice cloning - Covers 10 languages: English, German, French, and more - Latency as low as 97ms for real-time applications - Supports both streaming and non-streaming generation The model comes in two sizes (0.6B and 1.7B parameters), so you can pick based on your hardware and quality needs. Three modes to work with: 1. Custom Voice: Use pre-built premium voices with instruction-based style control 2. Voice Design: Describe the voice you want in plain English (or Chinese), and the model creates it 3. Voice Clone: Provide a 3-second reference audio and clone that voice The best part? It integrates with vLLM for production deployment and has a simple Python package you can pip install. I've shared a link to the GitHub repo in the next tweet.
0:17

Big moment for text-to-speech. Qwen just open-sourced a text-to-speech model that lets you clone voices, design new ones, and control speech using natural language. Let me explain what I mean: You can literally tell it "speak in a cheerful tone with slight nervousness," and it actually does that. No complex audio engineering needed. What makes this special: - 3-second voice cloning - Covers 10 languages: English, German, French, and more - Latency as low as 97ms for real-time applications - Supports both streaming and non-streaming generation The model comes in two sizes (0.6B and 1.7B parameters), so you can pick based on your hardware and quality needs. Three modes to work with: 1. Custom Voice: Use pre-built premium voices with instruction-based style control 2. Voice Design: Describe the voice you want in plain English (or Chinese), and the model creates it 3. Voice Clone: Provide a 3-second reference audio and clone that voice The best part? It integrates with vLLM for production deployment and has a simple Python package you can pip install. I've shared a link to the GitHub repo in the next tweet.

Akshay 🚀

31,249 views • 6 months ago

📢 EstateX: The Next Phase Begins This week, we're pulling back the curtain on what we've been building - and announcing the first wave of ecosystem activations. What's Coming This Week: ✅ Multiple Project updates ✅ Concrete dates for key launches ✅ Behind-the-scenes reveals Now that the launch is complete, it’s time for the REAL work and business building to begin. A series of exciting developments are scheduled for release in the near future, including but not limited to: 📊 The Ecosystem Overview: 👇 (All of which will have positive flow use cases and utilities for $ESX) 🔸 20-35% of Company Revenue Buybacks - this amount of revenue generated is used to buy the $ESX token off the market 🔸 Investment Platform Transaction Fees - all transaction fees on the EstateX investment platform are paid in $ESX or used to buy $ESX off the market 🔸 ESX Blockchain Gas Fees – Every transaction fuels the network, increasing $ESX usage and demand. 🔸 RWAPad / Unicorn Club Staking – Creating $ESX lockup and growth pressure. 🔸 Whitelabel Solutions – Platform adoption by other companies drives $ESX circulation and utility. 🔸 PropXChange – Trades in tokenized real estate generate fees that flow back to $ESX. 🔸 EstateX Pay – Card transactions generate revenue that supports $ESX growth. 🔸 CapitalX Lending – Loan and interest fees contribute to $ESX buybacks. 🔸 EstateX University – Access and education fees contribute to $ESX ecosystem growth. 🔸 $AIESX Utilization – Revenue drives $ESX buybacks. 🔸 Real Estate Rental Income – Passive rental flows create consistent $ESX circulation. 🔸 Expansion Into New RWAs – Future tokenized assets add more revenue streams, boosting $ESX demand. 🔸PLUS more surprises to be revealed! 💰 Financial Perspective: Consider just one of these components: real estate tokenization. A $100M AUM (Assets Under Management) portfolio in the first 6 months of operations could generate 10% in revenue from all pillars involved. That's $10M in total revenue. With up to 35% of that revenue allocated to buybacks, that's major buy pressure from a single revenue stream. Now multiply that across the entire ecosystem. But we're doing this RIGHT. That means rolling out strategically, testing thoroughly, and ensuring each piece is bulletproof before launch. 👉 Our new standard: Underpromise, overdeliver on EVERY deadline. This week, you'll see the first glimpses of what EstateX truly is. And this is just the beginning. First announcement drops very soon. The $ESX token is LIVE! Accumulate $ESX 👉 Watch the $ESX use case video now: Join our community 👇
0:12

📢 EstateX: The Next Phase Begins This week, we're pulling back the curtain on what we've been building - and announcing the first wave of ecosystem activations. What's Coming This Week: ✅ Multiple Project updates ✅ Concrete dates for key launches ✅ Behind-the-scenes reveals Now that the launch is complete, it’s time for the REAL work and business building to begin. A series of exciting developments are scheduled for release in the near future, including but not limited to: 📊 The Ecosystem Overview: 👇 (All of which will have positive flow use cases and utilities for $ESX) 🔸 20-35% of Company Revenue Buybacks - this amount of revenue generated is used to buy the $ESX token off the market 🔸 Investment Platform Transaction Fees - all transaction fees on the EstateX investment platform are paid in $ESX or used to buy $ESX off the market 🔸 ESX Blockchain Gas Fees – Every transaction fuels the network, increasing $ESX usage and demand. 🔸 RWAPad / Unicorn Club Staking – Creating $ESX lockup and growth pressure. 🔸 Whitelabel Solutions – Platform adoption by other companies drives $ESX circulation and utility. 🔸 PropXChange – Trades in tokenized real estate generate fees that flow back to $ESX. 🔸 EstateX Pay – Card transactions generate revenue that supports $ESX growth. 🔸 CapitalX Lending – Loan and interest fees contribute to $ESX buybacks. 🔸 EstateX University – Access and education fees contribute to $ESX ecosystem growth. 🔸 $AIESX Utilization – Revenue drives $ESX buybacks. 🔸 Real Estate Rental Income – Passive rental flows create consistent $ESX circulation. 🔸 Expansion Into New RWAs – Future tokenized assets add more revenue streams, boosting $ESX demand. 🔸PLUS more surprises to be revealed! 💰 Financial Perspective: Consider just one of these components: real estate tokenization. A $100M AUM (Assets Under Management) portfolio in the first 6 months of operations could generate 10% in revenue from all pillars involved. That's $10M in total revenue. With up to 35% of that revenue allocated to buybacks, that's major buy pressure from a single revenue stream. Now multiply that across the entire ecosystem. But we're doing this RIGHT. That means rolling out strategically, testing thoroughly, and ensuring each piece is bulletproof before launch. 👉 Our new standard: Underpromise, overdeliver on EVERY deadline. This week, you'll see the first glimpses of what EstateX truly is. And this is just the beginning. First announcement drops very soon. The $ESX token is LIVE! Accumulate $ESX 👉 Watch the $ESX use case video now: Join our community 👇

EstateX

118,901 views • 8 months ago

PLANES? ELON’S GOT ROCKETS FOR THAT Why sit for 15 hours when you could blast across Earth in 30 minutes at Mach 25?! Skip layovers - you'll be on time for dinner on the other side of the planet. The Rocket Ride Rundown: •⁠ ⁠New York to Sydney in under an hour - fasten your heat shields •⁠ ⁠Earth-to-Earth rockets - loud, proud, and ready to bounce borders •⁠ ⁠Teleport vibes - but with fire, noise, and a side of G-forces Boarding passes? Nah. Rocket rides don’t queue! Source: Technology
0:18

PLANES? ELON’S GOT ROCKETS FOR THAT Why sit for 15 hours when you could blast across Earth in 30 minutes at Mach 25?! Skip layovers - you'll be on time for dinner on the other side of the planet. The Rocket Ride Rundown: •⁠ ⁠New York to Sydney in under an hour - fasten your heat shields •⁠ ⁠Earth-to-Earth rockets - loud, proud, and ready to bounce borders •⁠ ⁠Teleport vibes - but with fire, noise, and a side of G-forces Boarding passes? Nah. Rocket rides don’t queue! Source: Technology

Mario Nawfal

44,798 views • 1 year ago

THE TESLA MODEL S: THE CAR THAT MADE ELECTRIC VEHICLES SERIOUS When the Model S launched in 2012, the entire world still saw EVs as slow, boring, short-range toys for tree-huggers. The Model S changed that narrative overnight. It wasn’t just an electric car — it was a statement. Here’s why the Model S was so important for EV adoption: • It proved EVs could be faster and better than gas cars 0–60 mph in under 4 seconds (later Plaid versions under 2 seconds) while being completely silent and smooth. It beat most supercars off the line and made “electric” synonymous with performance. • It delivered real long-range capability Over 300 miles of range when most EVs at the time struggled to reach 100 miles. Suddenly, road trips became possible and “range anxiety” started to feel outdated. • It introduced over-the-air updates The first production car that could get major performance upgrades, new features, and safety improvements wirelessly — like a smartphone on wheels. This changed how people think about car ownership forever. • It forced the entire auto industry to respond Legacy manufacturers who had been dragging their feet on EVs suddenly rushed to catch up. The Model S basically lit the fuse for the modern EV revolution. • It made luxury electric desirable Premium interior, massive touchscreen, ridiculous acceleration, and futuristic design turned EVs from “compromise” into “aspiration.” Without the Model S proving that electric cars could outperform and out-luxury gasoline vehicles, we wouldn’t have the Model 3/Y explosion, the Cybertruck, or the flood of competitors now racing to go electric. The Model S didn’t just sell cars. It changed the future of transportation. It took EVs from niche to mainstream and showed the world what was possible.
0:11

THE TESLA MODEL S: THE CAR THAT MADE ELECTRIC VEHICLES SERIOUS When the Model S launched in 2012, the entire world still saw EVs as slow, boring, short-range toys for tree-huggers. The Model S changed that narrative overnight. It wasn’t just an electric car — it was a statement. Here’s why the Model S was so important for EV adoption: • It proved EVs could be faster and better than gas cars 0–60 mph in under 4 seconds (later Plaid versions under 2 seconds) while being completely silent and smooth. It beat most supercars off the line and made “electric” synonymous with performance. • It delivered real long-range capability Over 300 miles of range when most EVs at the time struggled to reach 100 miles. Suddenly, road trips became possible and “range anxiety” started to feel outdated. • It introduced over-the-air updates The first production car that could get major performance upgrades, new features, and safety improvements wirelessly — like a smartphone on wheels. This changed how people think about car ownership forever. • It forced the entire auto industry to respond Legacy manufacturers who had been dragging their feet on EVs suddenly rushed to catch up. The Model S basically lit the fuse for the modern EV revolution. • It made luxury electric desirable Premium interior, massive touchscreen, ridiculous acceleration, and futuristic design turned EVs from “compromise” into “aspiration.” Without the Model S proving that electric cars could outperform and out-luxury gasoline vehicles, we wouldn’t have the Model 3/Y explosion, the Cybertruck, or the flood of competitors now racing to go electric. The Model S didn’t just sell cars. It changed the future of transportation. It took EVs from niche to mainstream and showed the world what was possible.

Tesla Owners Silicon Valley

11,056 views • 3 months ago

🚨🇷🇺 NATO IN PANIC: RUSSIA ARMS WORLD’S LARGEST CRUISER WITH SUBMARINE HUNTERS The Russian cruiser ‘Admiral Nakhimov’ is now ready to destroy any enemy submarine threat, equipped with three Ka-27M anti-submarine helicopters, making the world’s largest surface combatant the non-aircraft-carrier warship with the largest airborne squadron of any in service. 🔸 The 28,000-ton Admiral Nakhimov remains the world’s largest surface combatant and the only one designed from the outset to carry three heavyweight Ka-27M helicopters in a below-deck hangar. 🔸 The modernized Ka-27M helicopters deliver real-time digital sensor fusion, the Kopye-A radar for extended-range multi-target tracking, and advanced sonobuoy processing that better separates quiet submarine signatures from background noise. 🔸 These helicopters push detection hundreds of kilometers beyond hull-mounted sonars, laying active and passive sonobuoy fields while cueing the cruiser’s Otvet anti-submarine missiles against contacts the ship itself cannot yet hear. 🔸 In the Arctic and Norwegian Sea, Western submarines concentrate on monitoring and holding Russia’s sea-based nuclear deterrent at risk—this air wing addresses the growing difficulty of detecting stealthier boats from surface platforms alone. 🔸 Though the Kirov class excels in air defense and anti-ship strikes, its Arctic value may ultimately rest heaviest on these expanded helicopter-enabled anti-submarine capabilities. Do you think Admiral Nakhimov gives Russia a real anti-submarine edge in the Arctic?
0:13

🚨🇷🇺 NATO IN PANIC: RUSSIA ARMS WORLD’S LARGEST CRUISER WITH SUBMARINE HUNTERS The Russian cruiser ‘Admiral Nakhimov’ is now ready to destroy any enemy submarine threat, equipped with three Ka-27M anti-submarine helicopters, making the world’s largest surface combatant the non-aircraft-carrier warship with the largest airborne squadron of any in service. 🔸 The 28,000-ton Admiral Nakhimov remains the world’s largest surface combatant and the only one designed from the outset to carry three heavyweight Ka-27M helicopters in a below-deck hangar. 🔸 The modernized Ka-27M helicopters deliver real-time digital sensor fusion, the Kopye-A radar for extended-range multi-target tracking, and advanced sonobuoy processing that better separates quiet submarine signatures from background noise. 🔸 These helicopters push detection hundreds of kilometers beyond hull-mounted sonars, laying active and passive sonobuoy fields while cueing the cruiser’s Otvet anti-submarine missiles against contacts the ship itself cannot yet hear. 🔸 In the Arctic and Norwegian Sea, Western submarines concentrate on monitoring and holding Russia’s sea-based nuclear deterrent at risk—this air wing addresses the growing difficulty of detecting stealthier boats from surface platforms alone. 🔸 Though the Kirov class excels in air defense and anti-ship strikes, its Arctic value may ultimately rest heaviest on these expanded helicopter-enabled anti-submarine capabilities. Do you think Admiral Nakhimov gives Russia a real anti-submarine edge in the Arctic?

NewRulesGeopolitics

24,487 views • 24 days ago

Tesla cut its Gigacasting processing time from 180 seconds to 75 seconds — nearly 60% faster 🌊 The Model Y Juniper rear casting now weighs approximately 60 kg, down from 67 kg on the previous generation. Much of the industry conversation centers on press size and tonnage. The concrete gains in speed and mass on this high-volume part come from targeted process refinements inside the die and across the production system. -> Processing time reduced from 180s to 75s on the rear casting -> Part weight lowered by 7 kg through incremental design improvements -> Faster cycle achieved while improving microstructure and mechanical properties Conformal cooling makes the difference. Complex water channels, drilled or through 3D-printed inserts directly into the die steel, target hot spots and pull heat out rapidly and evenly across the entire casting. This accelerates solidification, reduces temperature gradients, and allows the part to be ejected sooner without defects. The result is a casting that is both lighter and stronger, produced in less than half the time. Supporting elements include advanced software that controls every injection parameter and runner/gate designs refined through five years of iteration since 2020 + close collaboration among casting designers, die engineers, production, and safety teams running high-volume lines on three continents. Tesla’s manufacturing edge is not the Giga Press hardware itself. It is the accumulated knowledge of how to run these machines at scale. 📊 The Gigacasting Database gives you the full picture of the market: Credit: Atomic Industries - Aaron Slodov ❌ Don't leave your insights to chance with the X algorithm ✅ Subscribe for free to my weekly newsletter about all things Gigacasting and magnesium Thixomolding: 📬
0:16

Tesla cut its Gigacasting processing time from 180 seconds to 75 seconds — nearly 60% faster 🌊 The Model Y Juniper rear casting now weighs approximately 60 kg, down from 67 kg on the previous generation. Much of the industry conversation centers on press size and tonnage. The concrete gains in speed and mass on this high-volume part come from targeted process refinements inside the die and across the production system. -> Processing time reduced from 180s to 75s on the rear casting -> Part weight lowered by 7 kg through incremental design improvements -> Faster cycle achieved while improving microstructure and mechanical properties Conformal cooling makes the difference. Complex water channels, drilled or through 3D-printed inserts directly into the die steel, target hot spots and pull heat out rapidly and evenly across the entire casting. This accelerates solidification, reduces temperature gradients, and allows the part to be ejected sooner without defects. The result is a casting that is both lighter and stronger, produced in less than half the time. Supporting elements include advanced software that controls every injection parameter and runner/gate designs refined through five years of iteration since 2020 + close collaboration among casting designers, die engineers, production, and safety teams running high-volume lines on three continents. Tesla’s manufacturing edge is not the Giga Press hardware itself. It is the accumulated knowledge of how to run these machines at scale. 📊 The Gigacasting Database gives you the full picture of the market: Credit: Atomic Industries - Aaron Slodov ❌ Don't leave your insights to chance with the X algorithm ✅ Subscribe for free to my weekly newsletter about all things Gigacasting and magnesium Thixomolding: 📬

Luca Greco

169,989 views • 1 month ago

🚨🇷🇺 ZELENSKY IN PANIC: Russia's New Defense System Makes NATO Drone Swarms Obsolete Rostec just dropped the ZAK-30 “Citadel” — Russia’s latest anti-drone system designed to shred NATO UAV tactics in real time. 🔸 The ZAK-30 Citadel packs a 30mm cannon that fires intelligent smart shells which automatically calculate the optimal detonation point based on the drone’s exact flight path, delivering lethal precision while dramatically conserving ammunition. 🔸 Each combat module carries a full 300-round load and can destroy aircraft-type drones at oblique ranges up to 1,000 meters and small multicopters at 800 meters, supported by integrated surveillance radar plus optical-electronic tracking. 🔸 The entire system operates with a remote command post positioned safely up to 1 km away from the firing modules and requires only 5 crew members for continuous 24/7 performance. 🔸 Already battle-proven and deployed in the special military operation zone, the Citadel has confirmed its high effectiveness against real-world Western-supplied drone attacks. Do you think NATO can catch up to Russian anti-drone technology?
0:18

🚨🇷🇺 ZELENSKY IN PANIC: Russia's New Defense System Makes NATO Drone Swarms Obsolete Rostec just dropped the ZAK-30 “Citadel” — Russia’s latest anti-drone system designed to shred NATO UAV tactics in real time. 🔸 The ZAK-30 Citadel packs a 30mm cannon that fires intelligent smart shells which automatically calculate the optimal detonation point based on the drone’s exact flight path, delivering lethal precision while dramatically conserving ammunition. 🔸 Each combat module carries a full 300-round load and can destroy aircraft-type drones at oblique ranges up to 1,000 meters and small multicopters at 800 meters, supported by integrated surveillance radar plus optical-electronic tracking. 🔸 The entire system operates with a remote command post positioned safely up to 1 km away from the firing modules and requires only 5 crew members for continuous 24/7 performance. 🔸 Already battle-proven and deployed in the special military operation zone, the Citadel has confirmed its high effectiveness against real-world Western-supplied drone attacks. Do you think NATO can catch up to Russian anti-drone technology?

NewRulesGeopolitics

107,535 views • 2 months ago

A viral paper "Language Model Represents Space and Time" recently claims that LLMs learn "world models". As much as I like Max Tegmark's works, I disagree with their definition of world model. World model is a core concept in AI agent and decision making. It is our mental simulation of how the world works given interventions (or lack thereof). A world model captures causality and intuitive physics, telling the agent what is likely and what is impossible. It can and should be used for counterfactual reasoning, i.e. "what ifs": what would happen if I knock over a cup of water? Where would I have been if I had not taken that bus? Yann LeCun Yann LeCun says it well in his position paper ( I quote: "Using such world models, animals can learn new skills with very few trials. They can predict the consequences of their actions, they can reason, plan, explore, and imagine new solutions to problems. Importantly, they can also avoid making dangerous mistakes when facing an unknown situation." The first use of the term World Model in deep policy learning is attributed to hardmaru & Jürgen Schmidhuber: In their seminal paper, an agent masters shooting skills in the popular game Doom (demo below) by learning in imagination, using an internal world model as a "physics simulator". To put in a simple Python math formula, world model learns a function F(s[0:t-1], a) -> s[t:], which takes as input the observed past and current action, and outputs plausible future states. Now the definition of World Model in Tegmark's paper seems to be about predicting GPS coordinates and time eras. I see this as just a classification task with no causal learning and simulation going on. You cannot make meaningful interventions against that model, nor can you optimize any decision making in a closed feedback loop. As for the "space & time neurons", I think they are most similar to the "sentiment neuron" that OpenAI published in 2017: Predicting GPS is conceptually no different from predicting sentiment in my opinion. I don't think their experimental results are wrong - just that their conclusion is on shaky grounds. I welcome any debate! Paper link:
0:15

A viral paper "Language Model Represents Space and Time" recently claims that LLMs learn "world models". As much as I like Max Tegmark's works, I disagree with their definition of world model. World model is a core concept in AI agent and decision making. It is our mental simulation of how the world works given interventions (or lack thereof). A world model captures causality and intuitive physics, telling the agent what is likely and what is impossible. It can and should be used for counterfactual reasoning, i.e. "what ifs": what would happen if I knock over a cup of water? Where would I have been if I had not taken that bus? Yann LeCun Yann LeCun says it well in his position paper ( I quote: "Using such world models, animals can learn new skills with very few trials. They can predict the consequences of their actions, they can reason, plan, explore, and imagine new solutions to problems. Importantly, they can also avoid making dangerous mistakes when facing an unknown situation." The first use of the term World Model in deep policy learning is attributed to hardmaru & Jürgen Schmidhuber: In their seminal paper, an agent masters shooting skills in the popular game Doom (demo below) by learning in imagination, using an internal world model as a "physics simulator". To put in a simple Python math formula, world model learns a function F(s[0:t-1], a) -> s[t:], which takes as input the observed past and current action, and outputs plausible future states. Now the definition of World Model in Tegmark's paper seems to be about predicting GPS coordinates and time eras. I see this as just a classification task with no causal learning and simulation going on. You cannot make meaningful interventions against that model, nor can you optimize any decision making in a closed feedback loop. As for the "space & time neurons", I think they are most similar to the "sentiment neuron" that OpenAI published in 2017: Predicting GPS is conceptually no different from predicting sentiment in my opinion. I don't think their experimental results are wrong - just that their conclusion is on shaky grounds. I welcome any debate! Paper link:

Jim Fan

594,014 views • 2 years ago

📢 Today marks a major milestone as Cat is one step closer to decentralizing $MEOW. Cat has officially set the sell tax for the token to 0% and renounced control over the contract. This means the token tax can no longer be altered, and Cat no longer has the ability to blacklist/whitelist addresses or pause token transfers. However, this also means that any funds sent to the contract address cannot be recovered, so exercise caution and avoid sending funds to $MEOW's token address. With that being said, let Cat leave you with some reasons on why you should hyper gamble on $MEOW 👇 🔸 We have held our position as the leading meme coin on ZKsync, both pre- and post-airdrop. 🔸 $MEOW never faded; it simply traded within a wide range for months, forming a solid bottom on the chart, followed by a steady trend of higher lows over several months. 🔸 The Zeek Army stood strong through a 90% dip. Cat's battle-hardened community never wavered, continuing to vibe and hang out even during the toughest times. 🔸 Despite a low market cap, Cat maintained significant mindshare on ZKsync. 🔸 $MEOW's market cap is currently low enough to have realistic potential for a 10x gain. 🔸 Wealth transfer happened. After 10 months of token distribution, all jeets are out. The supply is now held mostly by long-term believers. Before the $ZK airdrop, 25% of $MEOW was staked, and that number increased to 40% post-airdrop—(3,3) vibes. 🔸 We are not a Cabal coin. While a Cabal can fake trading volume and exchange listings, they can’t fake time. Cat has always been transparent and grinding alongside everyone in the trenches since day one. Now, imagine where $MEOW's price would be if ZKsync catches a narrative and hits the $2B market cap mark. There is no second best.
0:10

📢 Today marks a major milestone as Cat is one step closer to decentralizing $MEOW. Cat has officially set the sell tax for the token to 0% and renounced control over the contract. This means the token tax can no longer be altered, and Cat no longer has the ability to blacklist/whitelist addresses or pause token transfers. However, this also means that any funds sent to the contract address cannot be recovered, so exercise caution and avoid sending funds to $MEOW's token address. With that being said, let Cat leave you with some reasons on why you should hyper gamble on $MEOW 👇 🔸 We have held our position as the leading meme coin on ZKsync, both pre- and post-airdrop. 🔸 $MEOW never faded; it simply traded within a wide range for months, forming a solid bottom on the chart, followed by a steady trend of higher lows over several months. 🔸 The Zeek Army stood strong through a 90% dip. Cat's battle-hardened community never wavered, continuing to vibe and hang out even during the toughest times. 🔸 Despite a low market cap, Cat maintained significant mindshare on ZKsync. 🔸 $MEOW's market cap is currently low enough to have realistic potential for a 10x gain. 🔸 Wealth transfer happened. After 10 months of token distribution, all jeets are out. The supply is now held mostly by long-term believers. Before the $ZK airdrop, 25% of $MEOW was staked, and that number increased to 40% post-airdrop—(3,3) vibes. 🔸 We are not a Cabal coin. While a Cabal can fake trading volume and exchange listings, they can’t fake time. Cat has always been transparent and grinding alongside everyone in the trenches since day one. Now, imagine where $MEOW's price would be if ZKsync catches a narrative and hits the $2B market cap mark. There is no second best.

Zeek

42,353 views • 1 year ago

A large tank at an aerospace factory in Garden Grove started leaking. The chemical inside - methyl methacrylate, or MMA - is a clear liquid used to make airplane canopies and strong glues for aircraft parts. When air got in through the leak, the chemical began to harden and turn solid…just like superglue or epoxy does when you leave the tube open. This hardening process creates a lot of heat. With thousands of gallons reacting at once, the heat keeps building and speeds up the reaction even more. This “runaway” effect has raised the tank’s temperature, damaged the tank walls, and is releasing toxic fumes. Timeline • May 21, 2026 (Thursday, ~3:40 p.m.): The 34,000-gallon tank at the GKN Aerospace facility in Garden Grove develops a leak. Air reaches the roughly 7,000 gallons of MMA inside, starting an exothermic curing reaction. Orange County Fire Authority responders arrive on scene on Western Avenue. • May 21–22, 2026: Heat from the reaction builds quickly. The temperature of the material rises from normal room temperature to about 90 °F. Toxic vapors begin venting from the tank and internal pressure starts to increase. • May 22–24, 2026: Evacuation orders are issued and expanded for safety. Approximately 40,000 residents in Garden Grove and nearby areas are told to leave. GKN Aerospace chemists and local hazmat teams monitor the tank around the clock and work on cooling and containment steps. • As of May 25, 2026: The tank is still unstable. The runaway chemical reaction continues, and emergency crews remain on site performing damage-control operations. The liquid MMA itself is highly toxic. If the tank fails completely, thousands of gallons could spill into storm drains and waterways, causing serious long-term environmental damage. At the same time, the ongoing heat and gas buildup raise the danger of a sudden pressure-related rupture or explosion that could send a large cloud of toxic vapors over this heavily populated part of Southern California. The situation has not been brought under control.
0:10

A large tank at an aerospace factory in Garden Grove started leaking. The chemical inside - methyl methacrylate, or MMA - is a clear liquid used to make airplane canopies and strong glues for aircraft parts. When air got in through the leak, the chemical began to harden and turn solid…just like superglue or epoxy does when you leave the tube open. This hardening process creates a lot of heat. With thousands of gallons reacting at once, the heat keeps building and speeds up the reaction even more. This “runaway” effect has raised the tank’s temperature, damaged the tank walls, and is releasing toxic fumes. Timeline • May 21, 2026 (Thursday, ~3:40 p.m.): The 34,000-gallon tank at the GKN Aerospace facility in Garden Grove develops a leak. Air reaches the roughly 7,000 gallons of MMA inside, starting an exothermic curing reaction. Orange County Fire Authority responders arrive on scene on Western Avenue. • May 21–22, 2026: Heat from the reaction builds quickly. The temperature of the material rises from normal room temperature to about 90 °F. Toxic vapors begin venting from the tank and internal pressure starts to increase. • May 22–24, 2026: Evacuation orders are issued and expanded for safety. Approximately 40,000 residents in Garden Grove and nearby areas are told to leave. GKN Aerospace chemists and local hazmat teams monitor the tank around the clock and work on cooling and containment steps. • As of May 25, 2026: The tank is still unstable. The runaway chemical reaction continues, and emergency crews remain on site performing damage-control operations. The liquid MMA itself is highly toxic. If the tank fails completely, thousands of gallons could spill into storm drains and waterways, causing serious long-term environmental damage. At the same time, the ongoing heat and gas buildup raise the danger of a sudden pressure-related rupture or explosion that could send a large cloud of toxic vapors over this heavily populated part of Southern California. The situation has not been brought under control.

DesertThrottleDiaries

1,345,474 views • 2 months ago

A look back at our mission to do /more. 🔥🛠️ Months ago, we entered a new growth phase for xLaunchpad, driven by a simple commitment: to do more. More exciting startups, more platform improvements, and a deeper, more meaningful connection with our community. /new We’ve introduced multiple initiatives to make participation in xLaunchpad more accessible, while still keeping EGLD stakers at the core of the experience: 🔸 Challenges Portal – Built in partnership with ᕈulsar Money, this platform allows users to earn lottery tickets by completing simple tasks. 🔸 Creators Program – A reward system for users who create valuable content about xLaunchpad and its projects. 🔸 xLaunchpad Users – Loyal participants from previous launches now have a chance to earn lottery tickets. 🔸 Project Users – Projects launching on xLaunchpad can distribute lottery tickets to eligible users based on set criteria. 🔸 EGLD Holders – Tickets can now be earned simply by holding EGLD in a wallet. Beyond these initiatives, we’ve also strengthened relationships with key community members, empowering them to write educational content about upcoming startups with full creative control on their side. And of course, we’ve introduced more projects and focused on accelerating launch timelines. /better The platform and lottery system have undergone continuous improvements, such as making participation more accessible as detailed above: 🔸 Faster EGLD Reclaims - Users now get their EGLD back more quickly after participation. 🔸 Improved Communication – Blog & X Articles covering both general topics and project-specific updates. Increased community discussions and more structured feedback collection. 🔸 Website Enhancements - Various UI/UX tweaks, with /more to come 👀. 🔸 Integrating AshSwap - Users could buy lottery tickets for the last project with any tokens. 🔸 KYC Improvements - Streamlined processes, with /more to come 👀. Additionally, we’ve successfully co-launched projects with other launchpads and supported startups in securing partnerships, listings, business deals, and legal guidance. xLaunchpad remains one of the most compliant launchpads, ensuring projects align with MiCA regulations and broader industry standards. /next Looking ahead, our main focus remains on the product and the community. Every improvement is guided by real user interactions and feedback, ensuring xLaunchpad continues to evolve. At its core, our mission revolves around two key goals: 1️⃣ Bringing exciting startups to life 2️⃣ Providing users with investment opportunities on the best terms For xLaunchpad and its community to thrive together, it’s essential to recognize these two pillars while understanding the platform’s bigger purpose: making something possible that wasn’t before - allowing retail users to participate in the early launch of promising projects and benefit through multiple avenues. From investments to early adoption, and from unlocking key benefits to shaping young startups. We cannot stress enough how much effort the team puts into achieving these key goals, and we’re grateful to have you by our side on this mission. /more
0:10

A look back at our mission to do /more. 🔥🛠️ Months ago, we entered a new growth phase for xLaunchpad, driven by a simple commitment: to do more. More exciting startups, more platform improvements, and a deeper, more meaningful connection with our community. /new We’ve introduced multiple initiatives to make participation in xLaunchpad more accessible, while still keeping EGLD stakers at the core of the experience: 🔸 Challenges Portal – Built in partnership with ᕈulsar Money, this platform allows users to earn lottery tickets by completing simple tasks. 🔸 Creators Program – A reward system for users who create valuable content about xLaunchpad and its projects. 🔸 xLaunchpad Users – Loyal participants from previous launches now have a chance to earn lottery tickets. 🔸 Project Users – Projects launching on xLaunchpad can distribute lottery tickets to eligible users based on set criteria. 🔸 EGLD Holders – Tickets can now be earned simply by holding EGLD in a wallet. Beyond these initiatives, we’ve also strengthened relationships with key community members, empowering them to write educational content about upcoming startups with full creative control on their side. And of course, we’ve introduced more projects and focused on accelerating launch timelines. /better The platform and lottery system have undergone continuous improvements, such as making participation more accessible as detailed above: 🔸 Faster EGLD Reclaims - Users now get their EGLD back more quickly after participation. 🔸 Improved Communication – Blog & X Articles covering both general topics and project-specific updates. Increased community discussions and more structured feedback collection. 🔸 Website Enhancements - Various UI/UX tweaks, with /more to come 👀. 🔸 Integrating AshSwap - Users could buy lottery tickets for the last project with any tokens. 🔸 KYC Improvements - Streamlined processes, with /more to come 👀. Additionally, we’ve successfully co-launched projects with other launchpads and supported startups in securing partnerships, listings, business deals, and legal guidance. xLaunchpad remains one of the most compliant launchpads, ensuring projects align with MiCA regulations and broader industry standards. /next Looking ahead, our main focus remains on the product and the community. Every improvement is guided by real user interactions and feedback, ensuring xLaunchpad continues to evolve. At its core, our mission revolves around two key goals: 1️⃣ Bringing exciting startups to life 2️⃣ Providing users with investment opportunities on the best terms For xLaunchpad and its community to thrive together, it’s essential to recognize these two pillars while understanding the platform’s bigger purpose: making something possible that wasn’t before - allowing retail users to participate in the early launch of promising projects and benefit through multiple avenues. From investments to early adoption, and from unlocking key benefits to shaping young startups. We cannot stress enough how much effort the team puts into achieving these key goals, and we’re grateful to have you by our side on this mission. /more

xLaunchpad

22,919 views • 1 year ago

Wonderland: Navigating 3D Scenes from a Single Image Contributions: • First, we introduce a representation for controllable 3D generation by leveraging the generative priors from camera-guided video diffusion models. Unlike image models, video diffusion models are trained on extensive video datasets. This enables them to capture comprehensive spatial relationships within scenes across multiple views and embed a form of "3D awareness" in their latent space, which allows us to maintain 3D consistency in novel view synthesis. • Second, to achieve controllable novel view generation, we empower video models with precise control over specified camera motions. We introduce a novel dual-branch conditioning mechanism that effectively incorporates desired diverse camera trajectories into the video diffusion model. This enables expansion of a single image into a multi-view consistent capture of a 3D scene with precise pose control. • Third, to achieve efficient 3D reconstruction, we directly transform video latents into 3DGS. We propose a novel latent-based large reconstruction model (LaLRM) that lifts video latents to 3D in a feed-forward manner. With this design, during inference, our model directly predicts 3DGS from a single input image, effectively aligning the generation and reconstruction tasks—and bridging image space and 3D space—through the video latent space. Compared with reconstructing scenes from images, the video latent space offers a 256× spatial-temporal reduction while retaining essential and consistent 3D structural details. Such a high degree of compression is crucial, as it allows the LaLRM to handle a wider range of 3D scenes within the reconstruction framework, with the same memory constraints.
0:12

Wonderland: Navigating 3D Scenes from a Single Image Contributions: • First, we introduce a representation for controllable 3D generation by leveraging the generative priors from camera-guided video diffusion models. Unlike image models, video diffusion models are trained on extensive video datasets. This enables them to capture comprehensive spatial relationships within scenes across multiple views and embed a form of "3D awareness" in their latent space, which allows us to maintain 3D consistency in novel view synthesis. • Second, to achieve controllable novel view generation, we empower video models with precise control over specified camera motions. We introduce a novel dual-branch conditioning mechanism that effectively incorporates desired diverse camera trajectories into the video diffusion model. This enables expansion of a single image into a multi-view consistent capture of a 3D scene with precise pose control. • Third, to achieve efficient 3D reconstruction, we directly transform video latents into 3DGS. We propose a novel latent-based large reconstruction model (LaLRM) that lifts video latents to 3D in a feed-forward manner. With this design, during inference, our model directly predicts 3DGS from a single input image, effectively aligning the generation and reconstruction tasks—and bridging image space and 3D space—through the video latent space. Compared with reconstructing scenes from images, the video latent space offers a 256× spatial-temporal reduction while retaining essential and consistent 3D structural details. Such a high degree of compression is crucial, as it allows the LaLRM to handle a wider range of 3D scenes within the reconstruction framework, with the same memory constraints.

MrNeRF

52,849 views • 1 year ago

The term "continual learning" has become overloaded if you see it as an ML problem. One classic thread is about memorization: regularization-based continual learning methods, such as EWC, MAS, and SI, estimate which parameters mattered for previous tasks and resist changing them too much. One modern thread is about adaptation: test-time training and inference-time learning methods, such as TTT, adapt part of the model on the incoming test stream before making predictions. These are sometimes discussed as separate threads. But in modern scalable architectures, I think they are better seen as complementary constraints: a model that learns quickly at test time also benefits from a mechanism for deciding what not to forget. In our #ECCV2026 paper, we study this in large-scale 4D reconstruction: how to build fast spatial memory that can adapt over long observation streams while reducing collapse and forgetting. Instead of using fully plastic test-time updates, we stabilize fast-weight adaptation with an elastic prior that balances adaptation and memory. Key ideas: - Elastic Test-Time Training: Fisher-weighted consolidation for fast-weight updates - EMA anchor weights that provide a moving reference for stability - Chunk-by-chunk inference for long 3D/4D observation streams We show that this scales across large 3D/4D pretraining settings, including both LRM-style and LVSM-style models, and improves reconstruction across benchmarks including Stereo4D, NVIDIA, and DL3DV-140. We release model checkpoints across different design choices: resolution, post-training curriculum, and whether the model uses an explicit 4DGS intermediate representation. - Homepage: - Paper: - Code: - Models: This work is co-led with Xueyang Yu, contributed by Haoyu Zhen Yuncong Yang, and advised by Michigan SLED Lab Chuang Gan.
0:14

The term "continual learning" has become overloaded if you see it as an ML problem. One classic thread is about memorization: regularization-based continual learning methods, such as EWC, MAS, and SI, estimate which parameters mattered for previous tasks and resist changing them too much. One modern thread is about adaptation: test-time training and inference-time learning methods, such as TTT, adapt part of the model on the incoming test stream before making predictions. These are sometimes discussed as separate threads. But in modern scalable architectures, I think they are better seen as complementary constraints: a model that learns quickly at test time also benefits from a mechanism for deciding what not to forget. In our #ECCV2026 paper, we study this in large-scale 4D reconstruction: how to build fast spatial memory that can adapt over long observation streams while reducing collapse and forgetting. Instead of using fully plastic test-time updates, we stabilize fast-weight adaptation with an elastic prior that balances adaptation and memory. Key ideas: - Elastic Test-Time Training: Fisher-weighted consolidation for fast-weight updates - EMA anchor weights that provide a moving reference for stability - Chunk-by-chunk inference for long 3D/4D observation streams We show that this scales across large 3D/4D pretraining settings, including both LRM-style and LVSM-style models, and improves reconstruction across benchmarks including Stereo4D, NVIDIA, and DL3DV-140. We release model checkpoints across different design choices: resolution, post-training curriculum, and whether the model uses an explicit 4DGS intermediate representation. - Homepage: - Paper: - Code: - Models: This work is co-led with Xueyang Yu, contributed by Haoyu Zhen Yuncong Yang, and advised by Michigan SLED Lab Chuang Gan.

Martin Ziqiao Ma

33,047 views • 1 month ago

HOLY CRAP! I can't tell you how big this is for the medical community and drug discovery: Google Announces AlphaFold 3 AI. Details: Enhanced Molecular Prediction: AlphaFold 3 predicts the structure and interactions of all life's molecules, including proteins, DNA, RNA, ligands, and more, with unprecedented accuracy. Improved Interaction Accuracy: For protein interactions with other molecule types, AlphaFold 3 offers at least a 50% improvement over existing methods, and doubles the accuracy for some critical interactions. Transformative Potential for Science and Medicine: The model aims to deepen our understanding of biological processes and significantly advance drug discovery efforts. Accessibility for Researchers: AlphaFold 3's capabilities are largely accessible for free via the AlphaFold Server, providing an essential tool for scientific research. Drug Design Innovation: AlphaFold 3 is utilized by Isomorphic Labs in collaboration with pharmaceutical companies to accelerate drug design, potentially leading to new treatments for various diseases. Foundation in AlphaFold 2: Building on the breakthroughs of AlphaFold 2, this version extends its scope beyond proteins to a wide range of biomolecules, enhancing its utility in scientific research and application. Global Accessibility and Educational Support: The AlphaFold Server is a free platform for non-commercial research worldwide, supported by educational resources to foster wider adoption and innovation. Empowering Rapid Scientific Advancements: By making detailed molecular interactions easily accessible, AlphaFold 3 enables faster hypothesis testing and could reduce the time and cost typically associated with experimental protein-structure prediction. Responsible Development and Deployment: DeepMind has engaged with domain experts to assess the impacts and potential risks of AlphaFold, ensuring its responsible use in the scientific community. Broad Implications for Biology:AlphaFold 3 helps reveal complex cellular mechanisms and interactions, offering insights that could lead to improved agricultural crops, enhanced understanding of diseases, and novel therapeutic strategies.
0:12

HOLY CRAP! I can't tell you how big this is for the medical community and drug discovery: Google Announces AlphaFold 3 AI. Details: Enhanced Molecular Prediction: AlphaFold 3 predicts the structure and interactions of all life's molecules, including proteins, DNA, RNA, ligands, and more, with unprecedented accuracy. Improved Interaction Accuracy: For protein interactions with other molecule types, AlphaFold 3 offers at least a 50% improvement over existing methods, and doubles the accuracy for some critical interactions. Transformative Potential for Science and Medicine: The model aims to deepen our understanding of biological processes and significantly advance drug discovery efforts. Accessibility for Researchers: AlphaFold 3's capabilities are largely accessible for free via the AlphaFold Server, providing an essential tool for scientific research. Drug Design Innovation: AlphaFold 3 is utilized by Isomorphic Labs in collaboration with pharmaceutical companies to accelerate drug design, potentially leading to new treatments for various diseases. Foundation in AlphaFold 2: Building on the breakthroughs of AlphaFold 2, this version extends its scope beyond proteins to a wide range of biomolecules, enhancing its utility in scientific research and application. Global Accessibility and Educational Support: The AlphaFold Server is a free platform for non-commercial research worldwide, supported by educational resources to foster wider adoption and innovation. Empowering Rapid Scientific Advancements: By making detailed molecular interactions easily accessible, AlphaFold 3 enables faster hypothesis testing and could reduce the time and cost typically associated with experimental protein-structure prediction. Responsible Development and Deployment: DeepMind has engaged with domain experts to assess the impacts and potential risks of AlphaFold, ensuring its responsible use in the scientific community. Broad Implications for Biology:AlphaFold 3 helps reveal complex cellular mechanisms and interactions, offering insights that could lead to improved agricultural crops, enhanced understanding of diseases, and novel therapeutic strategies.

Brian Krassenstein

258,615 views • 2 years ago

Robotics keeps hitting the same wall. Single task RL works, but... it does not scale to hundreds of tasks or new embodiments. This new paper looks like a real step toward fixing that. The team introduces MMBench, a benchmark with 200 tasks across many domains and robots, and Newt, a language conditioned world model trained online across all 200 tasks at once. The simple idea behind Newt: The model learns from demos to get the right priors It trains across many tasks through online interaction It uses language to ground the goal It adapts fast when a new task shows up What stood out to me: ✅ One model trained on 200 tasks at the same time ✅ Language conditioned control for both states and RGB ✅ Better data efficiency than strong baselines ✅ Strong open loop control ✅ Fast adaptation to new tasks and embodiments ✅ Full release of 200 checkpoints, 4000 demos, code, and benchmark This is a good push toward general control instead of one model per task. If you want the full paper: Project page: —- Weekly robotics and AI insights. Subscribe free:
0:11

Robotics keeps hitting the same wall. Single task RL works, but... it does not scale to hundreds of tasks or new embodiments. This new paper looks like a real step toward fixing that. The team introduces MMBench, a benchmark with 200 tasks across many domains and robots, and Newt, a language conditioned world model trained online across all 200 tasks at once. The simple idea behind Newt: The model learns from demos to get the right priors It trains across many tasks through online interaction It uses language to ground the goal It adapts fast when a new task shows up What stood out to me: ✅ One model trained on 200 tasks at the same time ✅ Language conditioned control for both states and RGB ✅ Better data efficiency than strong baselines ✅ Strong open loop control ✅ Fast adaptation to new tasks and embodiments ✅ Full release of 200 checkpoints, 4000 demos, code, and benchmark This is a good push toward general control instead of one model per task. If you want the full paper: Project page: —- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

70,090 views • 8 months ago