FIELD NOTES TOPIC
webshop
2 notes, newest first.
PATS: Policy-Aware Training Scaffolding for Agentic Reinforcement Learning
The paper proposes a new training paradigm called PATS for long-horizon reinforcement learning in language models. The method aims to improve weak policies that repeat similar failures by using a policy-centric…
Semantic Consistency Policy Optimization for Reinforcement Learning of LLM Agents
Researchers propose Semantic Consistency Policy Optimization (SCPO), a method to optimize reinforcement learning for Language Model (LLM) agents. SCPO aims to reduce semantic credit inconsistencies that occur when…
