How I Study AI - Learn AI Papers & Lectures the Easy Way

THINKSAFE: Self-Generated Safety Alignment for Reasoning Models

Intermediate

Seanie Lee, Sangwoo Park et al.Jan 30arXiv

Large reasoning models got very good at thinking step-by-step, but that sometimes made them too eager to follow harmful instructions.

#THINKSAFE#self-generated safety alignment#refusal steering

Collaborative Multi-Agent Test-Time Reinforcement Learning for Reasoning

Intermediate

Zhiyuan Hu, Yunhai Hu et al.Jan 14arXiv

This paper introduces MATTRL, a way for multiple AI agents to learn from their own conversations at test time using short, reusable text notes instead of retraining their weights.

#multi-agent systems#test-time reinforcement learning#experience retrieval

Entropy Ratio Clipping as a Soft Global Constraint for Stable Reinforcement Learning

Intermediate

Zhenpeng Su, Leiyu Pan et al.Dec 5arXiv

Reinforcement learning (RL) can make big language models smarter, but off-policy training often pushes updates too far from the “safe zone,” causing unstable learning.

#reinforcement learning#PPO-clip#KL penalty

Papers3

THINKSAFE: Self-Generated Safety Alignment for Reasoning Models

Collaborative Multi-Agent Test-Time Reinforcement Learning for Reasoning

Entropy Ratio Clipping as a Soft Global Constraint for Stable Reinforcement Learning