Tootfinder

Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning
Shenzhi Wang, Le Yu, Chang Gao, Chujie Zheng, Shixuan Liu, Rui Lu, Kai Dang, Xionghui Chen, Jianxin Yang, Zhenru Zhang, Yuqiong Liu, An Yang, Andrew Zhao, Yang Yue, Shiji Song, Bowen Yu, Gao Huang, Junyang Lin
https://arx…

Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning
Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a powerful approach to enhancing the reasoning capabilities of Large Language Models (LLMs), while its mechanisms are not yet well understood. In this work, we undertake a pioneering exploration of RLVR through the novel perspective of token entropy patterns, comprehensively analyzing how different tokens influence reasoning performance. By examining token entropy patterns in Chain-of-Thought (CoT) reasoning, we observe that o…

CRScore : Reinforcement Learning with Verifiable Tool and AI Feedback for Code Review
Manav Nitin Kapadnis, Atharva Naik, Carolyn Rose
https://arxiv.org/abs/2506.00296

CRScore++: Reinforcement Learning with Verifiable Tool and AI Feedback for Code Review
Reinforcement learning (RL) to improve code review comment generation requires handling unstructured outputs, making reinforcement learning (RL) feedback challenging. The two main RL approaches, namely RL with Verifiable Feedback (RLVR) and RL with AI Feedback (RLAIF), offer trade-offs: RLVR provides reliable feedback for structured tasks like code generation, while RLAIF works for unstructured outputs but is subjective. We bridge this gap with CRScore++, an RL framework that leverages both LLM…

🇺🇦 #NowPlaying on BBCRadio3's #InTune
Charles Gounod, Royal Liverpool Philharmonic Orchestra & Libor Pešek:
🎵 Mors et Vita (Judex)
#CharlesGounod #RoyalLiverpoolPhilharmonicOrchestra #LiborPešek

Reasoning Like an Economist: Post-Training on Economic Problems Induces Strategic Generalization in LLMs
Yufa Zhou, Shaobo Wang, Xingyu Dong, Xiangqi Jin, Yifang Chen, Yue Min, Kexin Yang, Xingzhang Ren, Dayiheng Liu, Linfeng Zhang
https://arxiv.org/abs/2506.00577

Reasoning Like an Economist: Post-Training on Economic Problems Induces Strategic Generalization in LLMs
Directly training Large Language Models (LLMs) for Multi-Agent Systems (MAS) remains challenging due to intricate reward modeling, dynamic agent interactions, and demanding generalization requirements. This paper explores whether post-training techniques, specifically Supervised Fine-Tuning (SFT) and Reinforcement Learning with Verifiable Rewards (RLVR), can effectively $\textit{generalize}$ to multi-agent scenarios. We use economic reasoning as a testbed, leveraging its strong foundations in m…

🇺🇦 #NowPlaying on #BBC6Music's #HuwStephens
Beastie Boys:
🎵 (You Gotta) Fight for Your Right (To Party!)
#BeastieBoys
https://rolivaroliva.bandcamp.com/track/beastie-boys-fight-for-your-right-roliva-edit
https://open.spotify.com/track/5NLuC70kZQv8q34QyQa1DP

> She explained, “I just need the paper. I need to write things down. [The iPad script] also has so many passwords and then I would have ADD and then do something else and then it would lock me out … it was so complicated. So I printed it.”
When actors bypass corporate controls to get their work done, it's a showbiz story.

'Andor' Star Stellan Skarsgård Was the Ultimate 'Star Wars' Script Rulebreaker
Co-stars Diego Luna and Adria Arjona reminisced about the lengths the cast went to learn their lines on the highly secretive Disney+ show.

@… @… You don’t seem to be the only who has this association… I won’t comment, RLPers-VD oblige.

🇺🇦 #NowPlaying on BBCRadio3's #ComposerOfTheWeek #COTW
Dmitry Shostakovich, Royal Liverpool Philharmonic Orchestra & Vasily Petrenko:
🎵 Symphony No 7 "Leningrad" Op 60 (2nd mvt)
#DmitryShostakovich #RoyalLiverpoolPhilharmonicOrchestra #VasilyPetrenko

Stabilizing Knowledge, Promoting Reasoning: Dual-Token Constraints for RLVR
Jiakang Wang, Runze Liu, Fuzheng Zhang, Xiu Li, Guorui Zhou
https://arxiv.org/abs/2507.15778

Stabilizing Knowledge, Promoting Reasoning: Dual-Token Constraints for RLVR
Reinforcement Learning with Verifiable Rewards (RLVR) has become an effective post-training method for improving the reasoning abilities of Large Language Models (LLMs), mainly by shaping higher-order behaviors such as reflection and planning. However, previous RLVR algorithms often apply uniform training signals to all tokens, without considering the different roles of low-entropy knowledge-related tokens and high-entropy reasoning-related tokens. Some recent methods try to separate these toke…

🇺🇦 #NowPlaying on BBCRadio3's #ThisClassicalLife
Dani Howard, Peter Moore, Royal Liverpool Philharmonic Orchestra & Michael Seal:
🎵 Trombone Concerto (1st mvt. Realisation)
#DaniHoward #PeterMoore #RoyalLiverpoolPhilharmonicOrchestra #MichaelSeal