Turning Multi-Agent Communication into Token-Level Credit with On Policy Self Distillation
cs.AI
Submitted: 2026-11-12
Updated: 2026-11-12
Code: https://github.com/Camel-Prince/OPSD4MAS
Terminology
Sources
- Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs
- Improving Factuality and Reasoning in Language Models through Multiagent Debate
- Dr. MAS: Stable Reinforcement Learning for Multi-Agent LLM Systems
- Measuring Mathematical Problem Solving With the MATH Dataset
- Multi-Agent Deep Research: Training Multi-Agent Systems with M-GRPO
- EDGE-OPD: Internalizing Privileged Context with Evidence Guided On-Policy Distillation
- GEAR: Granularity-Adaptive Advantage Reweighting for LLM Agents via Self-Distillation
- Let's Verify Step by Step
- LLM Collaboration With Multi-Agent Reinforcement Learning
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- A Survey of On-Policy Distillation for Large Language Models
- AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
- Qwen3 Technical Report
- On-Policy Context Distillation for Language Models
- DAPO: An Open-Source LLM Reinforcement Learning System at Scale
- Multi-Rollout On-Policy Distillation via Peer Successes and Failures
- Beyond Absolute Imitation: Anchored Residual Guidance for Privileged On-Policy Distillation
- $\pi$-Play: Multi-Agent Self-Play via Privileged Self-Distillation without External Data
- MAS-OPD: On-Policy Distillation for Multi-agent Systems
- TurnOPD: Making On-Policy Distillation Turn-Aware for Efficient Long-Horizon Agent Training
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents