Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Can “chicken” make a base model reason? 🐔 Yes! Why? With the right first tokens, base models can match RL in reasoning. We trace this to learned training data associations and use a simple data edit to make “chicken” elicit reasoning too. blog:

Sophie Wang

1,979 subscribers

103,738 views • 1 day ago •via X (Twitter)

Education Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

We found a new way to get language models to reason. 🤯 No RL, no training, no verifiers, no prompting. ❌ With better sampling, base models can achieve single-shot reasoning on par with (or better than!) GRPO while avoiding its characteristic loss in generation diversity.

We found a new way to get language models to reason. 🤯 No RL, no training, no verifiers, no prompting. ❌ With better sampling, base models can achieve single-shot reasoning on par with (or better than!) GRPO while avoiding its characteristic loss in generation diversity.

Aayush Karan

280,576 views • 11 months ago

OpenAI's Deep Research is getting a run for its money. Deep Lake was just released, and it's a different take on an AI system that can do deep research on your own data. You can use Deep Lake to build AI search with reasoning on your private and public data. (Look at the attached videos to get an idea of how it works.) If you want to research proprietary and sensitive data, Deep Research won't help you because it's limited to public data. Deep Lake, however, will allow you to use your private data. On top of that, Deep Lake supports multi-modal retrieval from the ground up. It uses vision language models for data ingestion and retrieval so that you can connect any data (PDFs, images, videos, structured data, etc.) You can even use mixed-data queries! Deep Lake can search your data from S3, Dropbox, and GCP. It learns from your queries over time, making the results as relevant to your work as possible!

OpenAI's Deep Research is getting a run for its money. Deep Lake was just released, and it's a different take on an AI system that can do deep research on your own data. You can use Deep Lake to build AI search with reasoning on your private and public data. (Look at the attached videos to get an idea of how it works.) If you want to research proprietary and sensitive data, Deep Research won't help you because it's limited to public data. Deep Lake, however, will allow you to use your private data. On top of that, Deep Lake supports multi-modal retrieval from the ground up. It uses vision language models for data ingestion and retrieval so that you can connect any data (PDFs, images, videos, structured data, etc.) You can even use mixed-data queries! Deep Lake can search your data from S3, Dropbox, and GCP. It learns from your queries over time, making the results as relevant to your work as possible!

Santiago

171,340 views • 1 year ago

AI has had exactly two scaling axes that worked so far, and the second one is starting to look finite too the first one was pretraining: with scaling parameters and data, we got world knowledge (i.e. ChatGPT had read enough to know things), but it started saturating a while ago the second one was RL, and people had been doing RL the whole time before that: RLHF is RL but it never scaled far because it was trying to control the exact output, which tokens come out, how the text reads, but you can only push that so far before you’re just polishing RLVR dropped that constraint: giving the model a task, then checking whether the final answer is right, and ignoring everything in between -- so the model does whatever it wants in the middle and only the endpoint gets graded, and that’s much closer to actual RL and it’s what bought us planning and reasoning (arguably, tool use sits around 2.5 on this list -- while useful, it's not a different kind of thing) so one axis gave knowledge, the other gave reasoning, and both of them are one model working alone the next axis is how many models you can get working on the same problem, which is a different kind of axis than the previous two we know that multi-agent RL has always been the harder problem: I spent years in that literature and the gap between single-agent and multi-agent is definitely not incremental -- it’s a whole different class of difficulty! which is also why the derivatives are steep at the start, nobody has picked the easy wins yet... and the thing that gates this multi-agent coordination is communication: models can only coordinate as well as they can exchange information, and right now they do that by writing sentences to each other imagine what could we possibly achieve if we properly open that third axis development by letting models to exchange information in their native "language" without loosing any computational data that they produce during inference

AI has had exactly two scaling axes that worked so far, and the second one is starting to look finite too the first one was pretraining: with scaling parameters and data, we got world knowledge (i.e. ChatGPT had read enough to know things), but it started saturating a while ago the second one was RL, and people had been doing RL the whole time before that: RLHF is RL but it never scaled far because it was trying to control the exact output, which tokens come out, how the text reads, but you can only push that so far before you’re just polishing RLVR dropped that constraint: giving the model a task, then checking whether the final answer is right, and ignoring everything in between -- so the model does whatever it wants in the middle and only the endpoint gets graded, and that’s much closer to actual RL and it’s what bought us planning and reasoning (arguably, tool use sits around 2.5 on this list -- while useful, it's not a different kind of thing) so one axis gave knowledge, the other gave reasoning, and both of them are one model working alone the next axis is how many models you can get working on the same problem, which is a different kind of axis than the previous two we know that multi-agent RL has always been the harder problem: I spent years in that literature and the gap between single-agent and multi-agent is definitely not incremental -- it’s a whole different class of difficulty! which is also why the derivatives are steep at the start, nobody has picked the easy wins yet... and the thing that gates this multi-agent coordination is communication: models can only coordinate as well as they can exchange information, and right now they do that by writing sentences to each other imagine what could we possibly achieve if we properly open that third axis development by letting models to exchange information in their native "language" without loosing any computational data that they produce during inference

Sasha Malysheva

15,053 views • 1 month ago

All of Us has significantly expanded our data to now include nearly a quarter million whole genome sequences! Researchers can use this data to make scientific discoveries that can lead to a healthier future for everyone. Read the announcement #JoinAllofUs

All of Us has significantly expanded our data to now include nearly a quarter million whole genome sequences! Researchers can use this data to make scientific discoveries that can lead to a healthier future for everyone. Read the announcement #JoinAllofUs

AllofUsResearch

36,537 views • 3 years ago

Introducing RL Environment Creator Skill Now any one can create RL environments $ npx skills add adithya-s-k/RL_Envs_101 > You can create environments across multiple frameworks like OpenEnv, OpenReward, Verifiers, NemoGym ... > the repo has live working examples of environments that your coding agent can reference > The skill is design to first understand what type of model you are training and create an environment while keeping that in mind ps. There’s a lot more to building RL environments that can be used for training. One major aspect is the data, which this skill can’t directly solve. However, the skill will help with implementing tools, rewards, and other components of an RL environment, making it easier to go from idea to implementation quickly across different frameworks. Let me know if you’d be interested in a detailed, end-to-end blog/tutorial on building an environment and actually training a model for a useful use case.

Introducing RL Environment Creator Skill Now any one can create RL environments $ npx skills add adithya-s-k/RL_Envs_101 > You can create environments across multiple frameworks like OpenEnv, OpenReward, Verifiers, NemoGym ... > the repo has live working examples of environments that your coding agent can reference > The skill is design to first understand what type of model you are training and create an environment while keeping that in mind ps. There’s a lot more to building RL environments that can be used for training. One major aspect is the data, which this skill can’t directly solve. However, the skill will help with implementing tools, rewards, and other components of an RL environment, making it easier to go from idea to implementation quickly across different frameworks. Let me know if you’d be interested in a detailed, end-to-end blog/tutorial on building an environment and actually training a model for a useful use case.

Adithya S K

47,002 views • 4 months ago

How can we address the scarcity of data required for specialized AI? Learn about Simula, a framework that reframes synthetic data generation as dataset-level mechanism design. By using reasoning to architect datasets from first principles, Simula enables fine-grained control over coverage, complexity, and quality. More →

How can we address the scarcity of data required for specialized AI? Learn about Simula, a framework that reframes synthetic data generation as dataset-level mechanism design. By using reasoning to architect datasets from first principles, Simula enables fine-grained control over coverage, complexity, and quality. More →

Google Research

142,710 views • 5 months ago

☀️Very proud of the team for training 5.6! A few of my highlights: - Great front-end aesthetics. E.g., I asked Sol to make a new blog for 5.6 with a celestial theme. - Much better CUA - Can work for much longer and is less lazy But it's not all great: - The writing quality has improved, but it's still bad in artifacts - Too many options with models/reasoning/fast mode/Cerebras/multi-agent We'll fix the above. What else should we improve? PS: belated post as there are so many other good models to train.

☀️Very proud of the team for training 5.6! A few of my highlights: - Great front-end aesthetics. E.g., I asked Sol to make a new blog for 5.6 with a celestial theme. - Much better CUA - Can work for much longer and is less lazy But it's not all great: - The writing quality has improved, but it's still bad in artifacts - Too many options with models/reasoning/fast mode/Cerebras/multi-agent We'll fix the above. What else should we improve? PS: belated post as there are so many other good models to train.

Yann Dubois

37,253 views • 2 months ago

New research from Databricks: LLMs Can Learn to Reason via Off-Policy RL Optimal Advantage-based Policy Optimization with Lagged Inference policy (OAPL) shows you don’t need strict on-policy training to improve reasoning. It matches or beats Group Relative Policy Optimization (GRPO), stays stable with large policy lag, and uses ~3× fewer training generations. For Databricks customers, it’s a simpler, practical, and equally powerful approach to RL that Databricks is pioneering internally — and bringing directly to Databricks customers, so enterprises can improve agents using the same methods we use for our in-house agents, without complex infrastructure changes.

New research from Databricks: LLMs Can Learn to Reason via Off-Policy RL Optimal Advantage-based Policy Optimization with Lagged Inference policy (OAPL) shows you don’t need strict on-policy training to improve reasoning. It matches or beats Group Relative Policy Optimization (GRPO), stays stable with large policy lag, and uses ~3× fewer training generations. For Databricks customers, it’s a simpler, practical, and equally powerful approach to RL that Databricks is pioneering internally — and bringing directly to Databricks customers, so enterprises can improve agents using the same methods we use for our in-house agents, without complex infrastructure changes.

Databricks AI Research

12,823 views • 7 months ago

Google has just launched an AI feature capable of planning, reasoning and searching the web. All this is done in a single prompt using Deep Research. This is literally the first reasoning model that has access to the internet. Game-changer. How to use it below 🧵

Google has just launched an AI feature capable of planning, reasoning and searching the web. All this is done in a single prompt using Deep Research. This is literally the first reasoning model that has access to the internet. Game-changer. How to use it below 🧵

Paul Couvert

298,330 views • 1 year ago

We asked Claude to make a track on KORUS. KORUS MCP handled the creation. Base MCP handled the wallet and the payment. This is what AiFi looks like in practice. With a single prompt, you can now go from an idea to a licensed track you own, settled in seconds on Base via x402. Take a look 👇

We asked Claude to make a track on KORUS. KORUS MCP handled the creation. Base MCP handled the wallet and the payment. This is what AiFi looks like in practice. With a single prompt, you can now go from an idea to a licensed track you own, settled in seconds on Base via x402. Take a look 👇

KOR

21,586 views • 23 days ago

Partnership with Aerodrome + $EMP coming to base Base Exciting news! $EMP is now tradable on Base 🔵 We seeded liquidity on Aerodrome to make $EMP available to new holders on a new ecosystem: Holders can bridge $EMP tokens from Arbitrum / Ethereum to Base or Base to Ethereum using our official bridge: 🌉 Simulacrum AI already uses Base as its settlement layer, making it a natural next home for $EMP There's much more to come from $EMP on Base - stay tuned 👀

Partnership with Aerodrome + $EMP coming to base Base Exciting news! $EMP is now tradable on Base 🔵 We seeded liquidity on Aerodrome to make $EMP available to new holders on a new ecosystem: Holders can bridge $EMP tokens from Arbitrum / Ethereum to Base or Base to Ethereum using our official bridge: 🌉 Simulacrum AI already uses Base as its settlement layer, making it a natural next home for $EMP There's much more to come from $EMP on Base - stay tuned 👀

Arcana

62,393 views • 1 year ago

A simple outfit can feel sweeter with the right bag. The flower crochet design, beige base, and pouch shape make it cute for markets, brunch, vacations, and weekend walks. Pick your color and shop now.

A simple outfit can feel sweeter with the right bag. The flower crochet design, beige base, and pouch shape make it cute for markets, brunch, vacations, and weekend walks. Pick your color and shop now.

Floxnar

2,565,070 views • 3 months ago

Here's a little math problem about GRID 🤔 Multiple agentic capabilities available in one place Data sources across industries + Reasoning to bring it all together —————————————————————————— = The best answers you can ask for

Here's a little math problem about GRID 🤔 Multiple agentic capabilities available in one place Data sources across industries + Reasoning to bring it all together —————————————————————————— = The best answers you can ask for

Sentient

1,136,952 views • 1 year ago

it's crazy what a 1.5B model can do these days! "VibeThinker-1.5B is a 1.5-billion parameter dense language model. With a total training cost of only $7,800 USD, it achieves reasoning performance comparable to larger models like GPT OSS-20B Medium." runs perfectly on device!

it's crazy what a 1.5B model can do these days! "VibeThinker-1.5B is a 1.5-billion parameter dense language model. With a total training cost of only $7,800 USD, it achieves reasoning performance comparable to larger models like GPT OSS-20B Medium." runs perfectly on device!

Maziyar PANAHI

202,387 views • 10 months ago

GPT-5.5 by Reasoning Effort: I've asked it in Codex to create a physics-based visualisation of RL cycles for different sized models (70b, 1t, 10t), to demonstrate how the amount of RL you can do differs by model size. My assessment of each: - Low: weird slop - Medium: kinda cooked - High: sort of tried but ultimately incoherent - Extra High: elite - really nice idea and well executed Obviously this is just one shot, but worth trying different reasoning levels for the new models, medium seems to be pretty good for GPT-5.5 and it was really bad for many previous GPT models.

GPT-5.5 by Reasoning Effort: I've asked it in Codex to create a physics-based visualisation of RL cycles for different sized models (70b, 1t, 10t), to demonstrate how the amount of RL you can do differs by model size. My assessment of each: - Low: weird slop - Medium: kinda cooked - High: sort of tried but ultimately incoherent - Extra High: elite - really nice idea and well executed Obviously this is just one shot, but worth trying different reasoning levels for the new models, medium seems to be pretty good for GPT-5.5 and it was really bad for many previous GPT models.

Peter Gostev (SF: 22-26 June)

209,258 views • 5 months ago

Turns out you can train humanoid hands without any robot data. The idea in HUG is quite simple: (a) collect human data with smart glasses, (b) train a human manipulation model, (c) retarget to multi-fingered robot hands.

Turns out you can train humanoid hands without any robot data. The idea in HUG is quite simple: (a) collect human data with smart glasses, (b) train a human manipulation model, (c) retarget to multi-fingered robot hands.

Lerrel Pinto

34,942 views • 3 months ago