Rethinking Reinforcement Learning for Language Models: A New Approach
The Thinking Seeds framework introduces a fresh perspective on reinforcement learning for language models. By mixing on-policy and off-policy data at the token level, it enhances training stability…
Read the full articleYou might also wanna read
Reinforcement Learning to Train Large Language Models to Explain Human Decisions
Article URL: Comments URL: Points: 6 # Comments: 0
Applying Tree Search Techniques to Language Models: Lessons from AlphaZero and DeepSeek-R1
Personal website of Ayush Tambde
SPIRAL: A Reinforcement Learning Framework for Multi-Primitive Language Model Reasoning
Language model reasoning can be substantially improved at test time via scaffolds that scale inference compute across different primitives -
arxiv.org·13d ago
New Framework Formalizes Learning from Language Feedback with Provable Performance Guarantees
Interactively learning from observation and language feedback is an increasingly studied area driven by the emergence of large language mode
arxiv.org·29d ago
Bidirectional Evolutionary Search: A New Framework for Self-Improving Language Models
Search has been proposed as an effective method for self-improving language models and agentic systems, both for post-training sample genera
arxiv.org·1mo ago
LifeSkill: A Reinforcement Learning Framework for Online Lifelong Learning in LLM Agents
Lifelong learning is essential for Large Language Model (LLM) agents operating in dynamic, interactive environments. However, existing lifel
arxiv.org·16d ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.