Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

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,...

IlirAliu_'s profile picture

Ilir Aliu

50,507 subscribers

70,090 Aufrufe • vor 9 Monaten •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

🛠️ What if a robot could invent its own tools. And teach itself how to use them? That’s exactly what VLMgineer does: a new framework that lets Vision Language Models (VLMs) design physical tools and the actions to use them, entirely on their own. No templates. No human demonstrations. Just raw, AI-driven creativity. Why it matters ✅ Co-designs tools and actions together using VLMs, ensuring tight coupling between form and function ✅ Uses VLM-guided evolution (not random search) to refine designs intelligently ✅ Outperforms human-designed tools by +64.7% in task success across 12 RoboToolBench challenges ✅ Produces better-than-everyday tools for real manipulation tasks—measured in success rate and elegance It builds on the emerging trend of large-model-guided evolutionary design (like Eureka and AlphaEvolve) and brings it into physical robotics. It opens the door to general-purpose, automated hardware design, no strong priors needed. Code & paper: —- Weekly robotics and AI insights. Subscribe free:
0:18

🛠️ What if a robot could invent its own tools. And teach itself how to use them? That’s exactly what VLMgineer does: a new framework that lets Vision Language Models (VLMs) design physical tools and the actions to use them, entirely on their own. No templates. No human demonstrations. Just raw, AI-driven creativity. Why it matters ✅ Co-designs tools and actions together using VLMs, ensuring tight coupling between form and function ✅ Uses VLM-guided evolution (not random search) to refine designs intelligently ✅ Outperforms human-designed tools by +64.7% in task success across 12 RoboToolBench challenges ✅ Produces better-than-everyday tools for real manipulation tasks—measured in success rate and elegance It builds on the emerging trend of large-model-guided evolutionary design (like Eureka and AlphaEvolve) and brings it into physical robotics. It opens the door to general-purpose, automated hardware design, no strong priors needed. Code & paper: —- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

13,984 Aufrufe • vor 8 Monaten

🤖 Another zero-shot reward model is now in LeRobot: ROBOMETER. A general-purpose, zero-shot video-language reward model from University of South Carolina, UT Dallas, Massachusetts Institute of Technology (MIT), University of Washington, Ai2, and NVIDIA that predicts frame-level task progress. Trained on 1M+ trajectories from 21 robot embodiments, generalizes zero-shot to unseen tasks, scenes, and robots. 2.4–4.5x better downstream success rates across online RL, offline RL, data filtering, failure detection, and data retrieval for IL. Project: Paper:
0:14

🤖 Another zero-shot reward model is now in LeRobot: ROBOMETER. A general-purpose, zero-shot video-language reward model from University of South Carolina, UT Dallas, Massachusetts Institute of Technology (MIT), University of Washington, Ai2, and NVIDIA that predicts frame-level task progress. Trained on 1M+ trajectories from 21 robot embodiments, generalizes zero-shot to unseen tasks, scenes, and robots. 2.4–4.5x better downstream success rates across online RL, offline RL, data filtering, failure detection, and data retrieval for IL. Project: Paper:

LeRobot

32,625 Aufrufe • vor 3 Monaten

New Generation Model! 🚨 We're introducing the Mistral model to our expanding lineup of generation models. Mistral brings efficient performance and strong language understanding capabilities to our platform. Initial testing shows promising results in code comprehension and generation tasks, making it a valuable addition to development workflows. While we continue to optimize its implementation, early benchmarks demonstrate consistent and reliable outputs across various programming tasks.
0:17

New Generation Model! 🚨 We're introducing the Mistral model to our expanding lineup of generation models. Mistral brings efficient performance and strong language understanding capabilities to our platform. Initial testing shows promising results in code comprehension and generation tasks, making it a valuable addition to development workflows. While we continue to optimize its implementation, early benchmarks demonstrate consistent and reliable outputs across various programming tasks.

ALCHEMIST AI 🔮

16,435 Aufrufe • vor 1 Jahr

Trained on zero real-world data. Learned to walk, pick up boxes, and follow multi-step instructions... in the REAL world. ( 📌 Paper below) Researchers from Amazon FAR, Berkeley, Stanford, and CMU scanned real rooms with an iPhone, rebuilt them as 3D Gaussian Splatting scenes, then generated 48,000 synthetic trajectories of a Unitree G1 walking, grasping, and placing objects inside those virtual replicas. They rendered the robot's first-person camera view from each run and paired it with the matching language instruction and motion data. That's the dataset every humanoid team needs and nobody has: synced egocentric video + language + kinematics, at scale. Instead of collecting it in the real world, they manufactured it. They trained a vision-language-kinematics policy on that synthetic data alone, then deployed it on the physical G1 across five task types: navigation to a named object, lifting boxes of three different sizes with no per-size tuning, chained multi-step tasks, robustness to mid-task layout changes and flickering lights, and multi-minute long-horizon runs. No real-world fine-tuning at any point. Real-world interaction data has been the hard limit on humanoid learning... slow, expensive, and small. If scanning a room once and synthesizing thousands of labeled interactions holds up as a general recipe, that limit moves. Data stops being the bottleneck robotics teams have to solve for. 📌 Paper: Project: ——- Weekly robotics and AI insights. Subscribe free:
0:15

Trained on zero real-world data. Learned to walk, pick up boxes, and follow multi-step instructions... in the REAL world. ( 📌 Paper below) Researchers from Amazon FAR, Berkeley, Stanford, and CMU scanned real rooms with an iPhone, rebuilt them as 3D Gaussian Splatting scenes, then generated 48,000 synthetic trajectories of a Unitree G1 walking, grasping, and placing objects inside those virtual replicas. They rendered the robot's first-person camera view from each run and paired it with the matching language instruction and motion data. That's the dataset every humanoid team needs and nobody has: synced egocentric video + language + kinematics, at scale. Instead of collecting it in the real world, they manufactured it. They trained a vision-language-kinematics policy on that synthetic data alone, then deployed it on the physical G1 across five task types: navigation to a named object, lifting boxes of three different sizes with no per-size tuning, chained multi-step tasks, robustness to mid-task layout changes and flickering lights, and multi-minute long-horizon runs. No real-world fine-tuning at any point. Real-world interaction data has been the hard limit on humanoid learning... slow, expensive, and small. If scanning a room once and synthesizing thousands of labeled interactions holds up as a general recipe, that limit moves. Data stops being the bottleneck robotics teams have to solve for. 📌 Paper: Project: ——- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

12,950 Aufrufe • vor 1 Monat

Most of what I actually need help with, I never think to tell a model. But why is it on me to remember? Our new paper asks: what if AI could proactively specialize to individuals and the tasks they’re carrying out at this very moment? 🧵
0:15

Most of what I actually need help with, I never think to tell a model. But why is it on me to remember? Our new paper asks: what if AI could proactively specialize to individuals and the tasks they’re carrying out at this very moment? 🧵

Michelle Lam

50,390 Aufrufe • vor 4 Monaten

🚀Thrilled to share what we’ve been building at TRI over the past several months: our first Large Behavior Models (LBMs) are here! I’m proud to have been a core contributor to the multi-task policy learning and post-training efforts. At TRI, we’ve been researching how LBMs can help robots learn faster, better, and more efficiently. The key takeaways: ✅ We built an evaluation pipeline to benchmark LBM performance with real 𝐬𝐭𝐚𝐭𝐢𝐬𝐭𝐢𝐜𝐚𝐥 𝐜𝐨𝐧𝐟𝐢𝐝𝐞𝐧𝐜𝐞 ✅ Pre-training on hundreds of tasks makes models more robust—plus, we can teach new, complex tasks with 80% 𝐥𝐞𝐬𝐬 𝐝𝐚𝐭𝐚 ✅ The bigger and more diverse the pre-training, the better the results Check out our overview video, webpage and paper for more details: ✨ 🌎 📄 We hope this work helps move the field of robotics forward!
0:18

🚀Thrilled to share what we’ve been building at TRI over the past several months: our first Large Behavior Models (LBMs) are here! I’m proud to have been a core contributor to the multi-task policy learning and post-training efforts. At TRI, we’ve been researching how LBMs can help robots learn faster, better, and more efficiently. The key takeaways: ✅ We built an evaluation pipeline to benchmark LBM performance with real 𝐬𝐭𝐚𝐭𝐢𝐬𝐭𝐢𝐜𝐚𝐥 𝐜𝐨𝐧𝐟𝐢𝐝𝐞𝐧𝐜𝐞 ✅ Pre-training on hundreds of tasks makes models more robust—plus, we can teach new, complex tasks with 80% 𝐥𝐞𝐬𝐬 𝐝𝐚𝐭𝐚 ✅ The bigger and more diverse the pre-training, the better the results Check out our overview video, webpage and paper for more details: ✨ 🌎 📄 We hope this work helps move the field of robotics forward!

Zubair Irshad

20,377 Aufrufe • vor 1 Jahr

BREAKING 🚨: OpenAI is actively polishing its Tasks feature and there is a big chance we will see them announced today 👀 - Tasks Beta will allow users to schedule tasks like "send me AI news from TestingCatalog at 9 am" - These automations will be handled by a new model tool "jawbone" - There will be a new Notifications tab in settings, assumingly to control the way you will receive notifications about scheduled tasks Interestingly, the same feature is being in development for Gemini. What is the chance of seeing both of them released on the same day?
0:17

BREAKING 🚨: OpenAI is actively polishing its Tasks feature and there is a big chance we will see them announced today 👀 - Tasks Beta will allow users to schedule tasks like "send me AI news from TestingCatalog at 9 am" - These automations will be handled by a new model tool "jawbone" - There will be a new Notifications tab in settings, assumingly to control the way you will receive notifications about scheduled tasks Interestingly, the same feature is being in development for Gemini. What is the chance of seeing both of them released on the same day?

🚨 AI News | TestingCatalog

204,165 Aufrufe • vor 1 Jahr

Super excited about Hydra-0 from Hongyu Li and team! The key idea is to use flow as a shared visual interface across embodiments/objects for controllable video generation, allowing a single generalist world model to learn from human, handheld-gripper, and robot interaction data. My favorite result is the video below: start from a real video of a human doing the task (left), extract the desired object flow, and condition the model on that flow (right). The model then hallucinates a plausible robot motion that could produce the same object motion. Very cool glimpse of how a generalist world model can bridge human demonstrations and robot control.
0:13

Super excited about Hydra-0 from Hongyu Li and team! The key idea is to use flow as a shared visual interface across embodiments/objects for controllable video generation, allowing a single generalist world model to learn from human, handheld-gripper, and robot interaction data. My favorite result is the video below: start from a real video of a human doing the task (left), extract the desired object flow, and condition the model on that flow (right). The model then hallucinates a plausible robot motion that could produce the same object motion. Very cool glimpse of how a generalist world model can bridge human demonstrations and robot control.

Yunzhu Li

10,509 Aufrufe • vor 21 Tagen

Most imitation learning policies break when the camera moves or the robot changes. NOT THIS ONE 👇 [📍 Bookmark for later ] A new 3D scene representation encoder, tackles this by enabling zero-shot generalization to unseen embodiments and viewpoints… And it works with any IL algorithm. The trick? •Use a 2D foundation model to extract semantic features •Lift them into 3D space for localization (not semantics) •Condition the IL policy on this spatially grounded vector Across 93 simulated and 6 real tasks, Adapt3R: ✅ Maintains IL performance on LIBERO & MimicGen benchmarks ✅ Outperforms DP3 and 3D Diffuser Actor in most settings ✅ Holds >80% success on LIBERO even with large camera rotations Thanks for sharing this, Animesh Garg & Albert Wilcox! 📍Paper: Website: Code:
0:14

Most imitation learning policies break when the camera moves or the robot changes. NOT THIS ONE 👇 [📍 Bookmark for later ] A new 3D scene representation encoder, tackles this by enabling zero-shot generalization to unseen embodiments and viewpoints… And it works with any IL algorithm. The trick? •Use a 2D foundation model to extract semantic features •Lift them into 3D space for localization (not semantics) •Condition the IL policy on this spatially grounded vector Across 93 simulated and 6 real tasks, Adapt3R: ✅ Maintains IL performance on LIBERO & MimicGen benchmarks ✅ Outperforms DP3 and 3D Diffuser Actor in most settings ✅ Holds >80% success on LIBERO even with large camera rotations Thanks for sharing this, Animesh Garg & Albert Wilcox! 📍Paper: Website: Code:

Ilir Aliu

12,178 Aufrufe • vor 1 Jahr

Can robots learn without training❓ [𝗜𝘁'𝘀 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲𝗱 ⬇ ] Teaching robots to do complex tasks WITHOUT spending hours training them. Sounds cool, right? That's exactly what DIAL-MPC does! The first training-free method for whole-body torque control using full-order dynamics: ✅ Instantly checks if a robot's moves are right or wrong ✅ Adapts quickly to new tasks without needing extra training ✅ Could work hand-in-hand with other robot learning methods Robots are getting smarter AND faster without the need for long training sessions. Website: Paper: Code: Saw this first Haoru Xue 🙏
0:10

Can robots learn without training❓ [𝗜𝘁'𝘀 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲𝗱 ⬇ ] Teaching robots to do complex tasks WITHOUT spending hours training them. Sounds cool, right? That's exactly what DIAL-MPC does! The first training-free method for whole-body torque control using full-order dynamics: ✅ Instantly checks if a robot's moves are right or wrong ✅ Adapts quickly to new tasks without needing extra training ✅ Could work hand-in-hand with other robot learning methods Robots are getting smarter AND faster without the need for long training sessions. Website: Paper: Code: Saw this first Haoru Xue 🙏

Ilir Aliu

71,540 Aufrufe • vor 1 Jahr

If you’ve been ignoring Axis Robotics because it looks like another random points farm, read this. In the last 48 hours, Unitree said robotics is approaching its "ChatGPT moment", while the chairman of ACE Robotics believes it could happen by the end of 2027. But for robots to reach that level, they need a crazy amount of training data. That’s exactly what Axis is building. You control simulated robots directly from your browser, complete simple tasks and earn points. Those movements also help create data that can be used to train real robots. And this isn’t some tiny experiment anymore: - $12M raised - 123K+ contributors - 3M+ robot trajectories - Community data already used to train a real robot So yeah, we’re basically farming a potential airdrop while teaching our future robot servants how to work 😂 If you haven’t started yet, it’s completely free. You only need a tiny amount of gas on Base to sign your completed tasks. ✅ Start farming Axis points: Important: Sign every completed task from the History page, otherwise you won’t receive the points.
0:12

If you’ve been ignoring Axis Robotics because it looks like another random points farm, read this. In the last 48 hours, Unitree said robotics is approaching its "ChatGPT moment", while the chairman of ACE Robotics believes it could happen by the end of 2027. But for robots to reach that level, they need a crazy amount of training data. That’s exactly what Axis is building. You control simulated robots directly from your browser, complete simple tasks and earn points. Those movements also help create data that can be used to train real robots. And this isn’t some tiny experiment anymore: - $12M raised - 123K+ contributors - 3M+ robot trajectories - Community data already used to train a real robot So yeah, we’re basically farming a potential airdrop while teaching our future robot servants how to work 😂 If you haven’t started yet, it’s completely free. You only need a tiny amount of gas on Base to sign your completed tasks. ✅ Start farming Axis points: Important: Sign every completed task from the History page, otherwise you won’t receive the points.

Pranjal Bora 🧭

29,321 Aufrufe • vor 20 Tagen

New open-source agent harness just landed! I got early access to TrueForge by TrueFoundry and have been running it locally for the past few days. The harness layer deserves as much attention as the model, and open source matters here because you can inspect the loop, run it on your own infrastructure, and swap to the latest or cheaper models. TrueForge handles the runtime work that makes an agent reliable. It drives the tool-calling loop, manages context, coordinates subagents, and executes code in a sandbox, with any model you choose. Every tool call re-sends the growing context to the model, so in practice the harness controls most of what an agent costs to run. A few things stood out from my testing and their published benchmarks. Vendor-Neutral by design. It runs OpenAI, Anthropic, and Google models alongside open-weight models like Kimi, GLM, and DeepSeek. Model routing is a setting, and you can send each task to the model that fits it. On a 14-task enterprise agent benchmark, it matched the accuracy of Claude Managed Agents running the same Opus 4.8 model at roughly 30% lower cost per run (3.8M tokens vs 10M for the same answers). Routing the same tasks to GLM-5.2 held accuracy and brought cost down by about 75%, around $3 per run instead of $12. Fully self-hosted and Open Source (MIT License). I had it running locally with one command, with sandboxed code execution working out of the box. It's time to own your agent harness. Thanks to TrueFoundry for partnering on this post.
0:16

New open-source agent harness just landed! I got early access to TrueForge by TrueFoundry and have been running it locally for the past few days. The harness layer deserves as much attention as the model, and open source matters here because you can inspect the loop, run it on your own infrastructure, and swap to the latest or cheaper models. TrueForge handles the runtime work that makes an agent reliable. It drives the tool-calling loop, manages context, coordinates subagents, and executes code in a sandbox, with any model you choose. Every tool call re-sends the growing context to the model, so in practice the harness controls most of what an agent costs to run. A few things stood out from my testing and their published benchmarks. Vendor-Neutral by design. It runs OpenAI, Anthropic, and Google models alongside open-weight models like Kimi, GLM, and DeepSeek. Model routing is a setting, and you can send each task to the model that fits it. On a 14-task enterprise agent benchmark, it matched the accuracy of Claude Managed Agents running the same Opus 4.8 model at roughly 30% lower cost per run (3.8M tokens vs 10M for the same answers). Routing the same tasks to GLM-5.2 held accuracy and brought cost down by about 75%, around $3 per run instead of $12. Fully self-hosted and Open Source (MIT License). I had it running locally with one command, with sandboxed code execution working out of the box. It's time to own your agent harness. Thanks to TrueFoundry for partnering on this post.

elvis

11,303 Aufrufe • vor 23 Tagen

a moonshot engineer leaked the benchmark anthropic, openai and xai all buried the same week: kimi k3 beat opus 5, gpt-5.6 and grok 4.6 at $0.94 a task. stop paying anthropic $200 a month for opus 5 and openai $200 for gpt-5.6 when kimi does the same work for $8 the leak showed kimi k3 winning 9 of 12 categories against opus 5, gpt-5.6 and grok 4.6. within 48 hours all three labs quietly pushed pricing pages and one very specific comparison chart off their sites. nobody announced anything. they just deleted, which tells you everything the four numbers they scrubbed: cost per task · $0.94 vs $1.80 -> opus 5 charges $1.80 to finish one task. gpt-5.6 $1.04. grok 4.6 $0.61. kimi k3 $0.94 and it landed 487 of 500 clean -> anthropic is billing you double for a model that lost the benchmark it paid to promote the weights · free, sitting on huggingface right now -> the entire model is a public download. pull it, keep it, run it forever, nobody can switch it off -> a model you can hold cannot be rented at $200 a month. that single fact is what three labs deleted a chart over the switch · one line of bash -> moonshot ships an anthropic-compatible endpoint. one env variable and claude code points at kimi -> same cli, same keybindings, same /model. you change a url, opus 5 never knows it lost the seat the bill · $400 down to $8 -> opus 5 max plus gpt-5.6 pro is $400 a month. kimi runs the same daily work for $8 metered -> that is a 98% cut for output that beat both of them 9 categories to 3 here is the part they will fight me on: the frontier tax died the week this leaked and all three labs know it. once the weights are public the price has a ceiling, because anyone can serve the same model. anthropic, openai and xai are charging 2025 prices on a lead that ended in a benchmark they deleted instead of answered drop your $400/mo ai stack to $8. the run above is kimi k3 finishing the task opus 5 bills $1.80 for. the full breakdown is in the article below
0:19

a moonshot engineer leaked the benchmark anthropic, openai and xai all buried the same week: kimi k3 beat opus 5, gpt-5.6 and grok 4.6 at $0.94 a task. stop paying anthropic $200 a month for opus 5 and openai $200 for gpt-5.6 when kimi does the same work for $8 the leak showed kimi k3 winning 9 of 12 categories against opus 5, gpt-5.6 and grok 4.6. within 48 hours all three labs quietly pushed pricing pages and one very specific comparison chart off their sites. nobody announced anything. they just deleted, which tells you everything the four numbers they scrubbed: cost per task · $0.94 vs $1.80 -> opus 5 charges $1.80 to finish one task. gpt-5.6 $1.04. grok 4.6 $0.61. kimi k3 $0.94 and it landed 487 of 500 clean -> anthropic is billing you double for a model that lost the benchmark it paid to promote the weights · free, sitting on huggingface right now -> the entire model is a public download. pull it, keep it, run it forever, nobody can switch it off -> a model you can hold cannot be rented at $200 a month. that single fact is what three labs deleted a chart over the switch · one line of bash -> moonshot ships an anthropic-compatible endpoint. one env variable and claude code points at kimi -> same cli, same keybindings, same /model. you change a url, opus 5 never knows it lost the seat the bill · $400 down to $8 -> opus 5 max plus gpt-5.6 pro is $400 a month. kimi runs the same daily work for $8 metered -> that is a 98% cut for output that beat both of them 9 categories to 3 here is the part they will fight me on: the frontier tax died the week this leaked and all three labs know it. once the weights are public the price has a ceiling, because anyone can serve the same model. anthropic, openai and xai are charging 2025 prices on a lead that ended in a benchmark they deleted instead of answered drop your $400/mo ai stack to $8. the run above is kimi k3 finishing the task opus 5 bills $1.80 for. the full breakdown is in the article below

starmex

32,547 Aufrufe • vor 21 Tagen

SOMEONE TURNED THEIR TEAM'S TASK TRACKER INTO A 3D ISLAND instead of a boring list of tasks, your teams work is a little island that grows as you get stuff done > you assign tasks right in slack, just type who its for, the points, and the due date > finish a task and you get to place a building on the island > get your work rejected and the building collapses into rubble > the rubble stays there forever, so everyone can see it > each new sprint starts a fresh island so over time the island fills up with buildings for all the work your team actually finished, and the rubble is a reminder of what got rejected. its open source, so any team can set it up. way more fun than staring at a to do list all day
0:13

SOMEONE TURNED THEIR TEAM'S TASK TRACKER INTO A 3D ISLAND instead of a boring list of tasks, your teams work is a little island that grows as you get stuff done > you assign tasks right in slack, just type who its for, the points, and the due date > finish a task and you get to place a building on the island > get your work rejected and the building collapses into rubble > the rubble stays there forever, so everyone can see it > each new sprint starts a fresh island so over time the island fills up with buildings for all the work your team actually finished, and the rubble is a reminder of what got rejected. its open source, so any team can set it up. way more fun than staring at a to do list all day

Om Patel

12,978 Aufrufe • vor 2 Monaten

AI in robotics gets all the attention right now, but sometimes the most interesting work is very practical. Viet built a small vision system that counts potatoes on a conveyor belt. No giant dataset. No huge model. Just a clear problem and a smart setup. He used Ultralytics’ ObjectCounter, trained a tiny YOLO11 nano model, and because there was no potato dataset, he annotated a single frame with SAM 2 and trained from that. One frame. Still works across the whole video. It is a good reminder that useful AI in industry often looks like this. Focused. Lightweight. Solves a real task. If you work in manufacturing or robotics, these small systems are usually the fastest wins. They save time, reduce errors, and do not need massive infrastructure. Nice work, Viet. His projects: —- Weekly robotics and AI insights. Subscribe free:
0:13

AI in robotics gets all the attention right now, but sometimes the most interesting work is very practical. Viet built a small vision system that counts potatoes on a conveyor belt. No giant dataset. No huge model. Just a clear problem and a smart setup. He used Ultralytics’ ObjectCounter, trained a tiny YOLO11 nano model, and because there was no potato dataset, he annotated a single frame with SAM 2 and trained from that. One frame. Still works across the whole video. It is a good reminder that useful AI in industry often looks like this. Focused. Lightweight. Solves a real task. If you work in manufacturing or robotics, these small systems are usually the fastest wins. They save time, reduce errors, and do not need massive infrastructure. Nice work, Viet. His projects: —- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

1,676,995 Aufrufe • vor 9 Monaten

Don't train the model, evolve the harness. I read a brilliant blog post from Hugging Face where they took a frozen open model scoring 0% on a hard legal agent benchmark, left its weights alone, and let an automated loop rewrite only the code around it. That code layer is the harness, the runtime wrapper that feeds the model context, runs its tool calls, and decides when a run ends. By the time the loop finished, the system had essentially matched Sonnet 4.6 on the benchmark's headline metric, at roughly 7x lower cost per task. Zero weights changed. The gain existed because of where the model was failing. The judge only grades files saved in the right place under the exact requested filename, and the model kept doing the legal analysis correctly, then saving it under the wrong name, dropping it in a scratch folder, or never writing it at all. So the 0% was never measuring legal reasoning. It was measuring the harness. Hand-tuning that layer is slow and model-specific, so they automated it. A Claude proposer adds exactly one mechanism per iteration, and an outer loop keeps it only if it clearly beats the current best, so accepted mechanisms compound. What the loop discovered says a lot about where agents actually fail. → The biggest single gain was file handling, not intelligence. An automatic step that lands the deliverable exactly where the judge expects it beat every prompt change, with zero extra model tokens. → Code fixes transferred across models, prompt playbooks did not. The same harness lifted a smaller model from the same family by 14 points, but the tuned prompts hurt a different model family on tasks it could already finish. → The harness mattered more than anything else. Same model, same judge, same tasks, and five different harnesses scored anywhere between 3.5% and 80.1%. The gains do eventually flatten, and the remaining misses look like real capability gaps. At some point the wrapper runs out of tricks and the model has to carry the work. But the lesson holds. A benchmark score measures the model and its harness together, and until the harness is fixed, it's impossible to know which one failed. I highly recommend reading this: I also wrote a deep dive on agent harness engineering a while back, covering the orchestration loop, tools, memory, context management, and everything that turns a stateless LLM into a capable agent. The article is quoted below.
0:18

Don't train the model, evolve the harness. I read a brilliant blog post from Hugging Face where they took a frozen open model scoring 0% on a hard legal agent benchmark, left its weights alone, and let an automated loop rewrite only the code around it. That code layer is the harness, the runtime wrapper that feeds the model context, runs its tool calls, and decides when a run ends. By the time the loop finished, the system had essentially matched Sonnet 4.6 on the benchmark's headline metric, at roughly 7x lower cost per task. Zero weights changed. The gain existed because of where the model was failing. The judge only grades files saved in the right place under the exact requested filename, and the model kept doing the legal analysis correctly, then saving it under the wrong name, dropping it in a scratch folder, or never writing it at all. So the 0% was never measuring legal reasoning. It was measuring the harness. Hand-tuning that layer is slow and model-specific, so they automated it. A Claude proposer adds exactly one mechanism per iteration, and an outer loop keeps it only if it clearly beats the current best, so accepted mechanisms compound. What the loop discovered says a lot about where agents actually fail. → The biggest single gain was file handling, not intelligence. An automatic step that lands the deliverable exactly where the judge expects it beat every prompt change, with zero extra model tokens. → Code fixes transferred across models, prompt playbooks did not. The same harness lifted a smaller model from the same family by 14 points, but the tuned prompts hurt a different model family on tasks it could already finish. → The harness mattered more than anything else. Same model, same judge, same tasks, and five different harnesses scored anywhere between 3.5% and 80.1%. The gains do eventually flatten, and the remaining misses look like real capability gaps. At some point the wrapper runs out of tricks and the model has to carry the work. But the lesson holds. A benchmark score measures the model and its harness together, and until the harness is fixed, it's impossible to know which one failed. I highly recommend reading this: I also wrote a deep dive on agent harness engineering a while back, covering the orchestration loop, tools, memory, context management, and everything that turns a stateless LLM into a capable agent. The article is quoted below.

Akshay 🚀

244,990 Aufrufe • vor 2 Monaten

Multi-robot learning is getting a serious boost! 📚 Researchers have extended Isaac Lab to train heterogeneous multi-agent robotic policies at scale. The new framework supports high-resolution physics, GPU-accelerated simulation, and both homogeneous and heterogeneous agents working together on coordination tasks. They benchmarked different approaches (MAPPO: Multi-Agent Proximal Policy Optimization and HAPPO: Heterogeneous Agent PPO) across six challenging scenarios and showed that large-scale multi-robot training is not only feasible, but efficient. It’s an important step for real-world robotic collaboration, where teams of robots need to coordinate, split tasks, adapt roles, and interact dynamically, not just operate as identical clones. The code is open-source, and it pushes Isaac Lab closer to what robotics actually needs: scalable, physics-driven environments where many different robots can learn to work together. Here's the project page: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →
0:16

Multi-robot learning is getting a serious boost! 📚 Researchers have extended Isaac Lab to train heterogeneous multi-agent robotic policies at scale. The new framework supports high-resolution physics, GPU-accelerated simulation, and both homogeneous and heterogeneous agents working together on coordination tasks. They benchmarked different approaches (MAPPO: Multi-Agent Proximal Policy Optimization and HAPPO: Heterogeneous Agent PPO) across six challenging scenarios and showed that large-scale multi-robot training is not only feasible, but efficient. It’s an important step for real-world robotic collaboration, where teams of robots need to coordinate, split tasks, adapt roles, and interact dynamically, not just operate as identical clones. The code is open-source, and it pushes Isaac Lab closer to what robotics actually needs: scalable, physics-driven environments where many different robots can learn to work together. Here's the project page: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

38,997 Aufrufe • vor 9 Monaten

DAO Labs Sneak Peak Preview: 1 ) Instant Sign Up/Sign across all HUBs via X or Wallet✅ 2 ) Profile summarized Data of your activities and what they are worth, get a better view of your earnings.💰 3 ) Task Navigator to oversee, in real time, what tasks are available for you to work on across all our HUBs.🧭 4 ) A Timer Function, being able to optimize post relevance and expiration.⏰ Half of all the features complete, on the way to grant you a seamless #SocialMining experience connecting all HUBs
0:16

DAO Labs Sneak Peak Preview: 1 ) Instant Sign Up/Sign across all HUBs via X or Wallet✅ 2 ) Profile summarized Data of your activities and what they are worth, get a better view of your earnings.💰 3 ) Task Navigator to oversee, in real time, what tasks are available for you to work on across all our HUBs.🧭 4 ) A Timer Function, being able to optimize post relevance and expiration.⏰ Half of all the features complete, on the way to grant you a seamless #SocialMining experience connecting all HUBs

DAO Labs

107,813 Aufrufe • vor 2 Jahren