Stabilizing On-Policy Targeted Guidance in Agentic Reinforcement Learning
cs.LG
Submitted: 2026-11-03
Updated: 2026-11-04
Code: https://github.com/Dao-AILab/quack
Terminology
Sources
- What Matters In On-Policy Reinforcement Learning? A Large-Scale Empirical Study
- A General Language Assistant as a Laboratory for Alignment
- Language Models are Hidden Reasoners: Unlocking Latent Reasoning Capabilities via Self-Rewarding
- CDE: Curiosity-Driven Exploration for Efficient Reinforcement Learning in Large Language Models
- HDPO: Hybrid Distillation Policy Optimization via Privileged Self-Distillation
- Implementation Matters in Deep Policy Gradients: A Case Study on PPO and TRPO
- Long-Horizon Language Model Reinforcement Learning via Progressive Point Matching
- Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRs
- Navigate the Unknown: Enhancing LLM Reasoning with Intrinsic Motivation Guided Exploration
- Variational Intrinsic Control
- Distilling the Knowledge in a Neural Network
- Reinforcement Learning via Self-Distillation
- Entropy-Aware On-Policy Distillation of Language Models
- Rethinking On-Policy Self-Distillation for Thinking Models
- Why Does Self-Distillation (Sometimes) Degrade the Reasoning Capability of LLMs?
- Efficient Exploration via State Marginal Matching
- Unifying Group-Relative and Self-Distillation Policy Optimization via Sample Routing
- Jointly Reinforcing Diversity and Quality in Language Model Generations
- Self-Distilled Agentic Reinforcement Learning
- Adaptive Guidance Accelerates Reinforcement Learning of Reasoning Models
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis