Natasha Jaques
@natashajaques.bsky.social
Assistant Professor at UW and Staff Research Scientist at Google DeepMind. Social Reinforcement Learning in multi-agent and human-AI interactions. PhD from MIT. Check out https://socialrl.cs.washington.edu/ and https://natashajaques.ai/.
By optimizing for intrinsic curiosity, the LLM learns how to ask a series of questions over the course of the conversation to improve the accuracy of its user model. This generates conversations which reveal significantly more information about the user.
July 8, 2025 at 11:06 PM
By optimizing for intrinsic curiosity, the LLM learns how to ask a series of questions over the course of the conversation to improve the accuracy of its user model. This generates conversations which reveal significantly more information about the user.
This work shows the benefit of RL training for improving reasoning skills when there is no possibility for data leakage. AND how continuously evolving multi-agent competition leads to the development of emergent skills that generalize to novel tasks.
July 1, 2025 at 8:24 PM
This work shows the benefit of RL training for improving reasoning skills when there is no possibility for data leakage. AND how continuously evolving multi-agent competition leads to the development of emergent skills that generalize to novel tasks.
We analyze the results and find that LLMs learn emergent reasoning patterns like case-by-case analysis and expected value calculation that transfer to improve performance on math questions.
July 1, 2025 at 8:24 PM
We analyze the results and find that LLMs learn emergent reasoning patterns like case-by-case analysis and expected value calculation that transfer to improve performance on math questions.
Way to go KJ for producing such an insightful paper in the first few months of your PhD!
April 19, 2025 at 5:03 AM
Way to go KJ for producing such an insightful paper in the first few months of your PhD!
I had a ton of fun using this as a kid. I actually made my high school English class project a giant hypercard-based video game where I drew each frame in Paint and hid buttons behind the hand-drawn elements that let you navigate the world. That was so fun...😍
March 28, 2025 at 7:57 PM
I had a ton of fun using this as a kid. I actually made my high school English class project a giant hypercard-based video game where I drew each frame in Paint and hid buttons behind the hand-drawn elements that let you navigate the world. That was so fun...😍
Yes! Or you could focus on developing better MARL algorithms for the corporations that let them cooperate to solve the social dilemma more effectively. Similar to MARL benchmarks like Melting Pot but for a more impactful domain
February 16, 2025 at 11:27 PM
Yes! Or you could focus on developing better MARL algorithms for the corporations that let them cooperate to solve the social dilemma more effectively. Similar to MARL benchmarks like Melting Pot but for a more impactful domain
In contrast, MARL enables testing new policies with many more agents over a long time horizon.
We hope this benchmark will enable researchers in the RL and MARL communities to develop sophisticated cooperation algorithms in the context of a societally impactful problem!
We hope this benchmark will enable researchers in the RL and MARL communities to develop sophisticated cooperation algorithms in the context of a societally impactful problem!
February 13, 2025 at 6:42 AM
In contrast, MARL enables testing new policies with many more agents over a long time horizon.
We hope this benchmark will enable researchers in the RL and MARL communities to develop sophisticated cooperation algorithms in the context of a societally impactful problem!
We hope this benchmark will enable researchers in the RL and MARL communities to develop sophisticated cooperation algorithms in the context of a societally impactful problem!
I’m really excited about this, as I think MARL provides a new tool in the toolbox for investigating this problem. Existing work on ESG disclosures focuses on empirical studies (can’t test counterfactual policies), or analytical economics models (limited to 2 players or short time intervals)
February 13, 2025 at 6:41 AM
I’m really excited about this, as I think MARL provides a new tool in the toolbox for investigating this problem. Existing work on ESG disclosures focuses on empirical studies (can’t test counterfactual policies), or analytical economics models (limited to 2 players or short time intervals)
...providing corporations with more reliable information about climate risks — and we show that this significantly improves corporations’ ability to mitigate climate change, even without the influence of investors!
February 13, 2025 at 6:39 AM
...providing corporations with more reliable information about climate risks — and we show that this significantly improves corporations’ ability to mitigate climate change, even without the influence of investors!