Revisiting On-Policy Distillation: Empirical Failure Modes and Simple Fixes
Revisiting On-Policy Distillation: Empirical Failure Modes and Simple Fixes explores Improves on-policy distillation for LLMs by addressing empirical failure modes to achieve more stable optimization and better downstream performance.. Commercial viability score: 4/10 in LLM Training.
BUILDER'S SANDBOX
Build This Paper
Use an AI coding agent to implement this research.
OpenAI CodexAI Agent
Lightweight coding agent in your terminal.
Claude CodeAI Agent
Agentic coding tool for terminal workflows.
AntiGravity IDEScaffolding
AI agent mindset installer and workflow scaffolder.
CursorIDE
AI-first code editor built on VS Code.
VS CodeIDE
Free, open-source editor by Microsoft.
Recommended Stack
Startup Essentials
Estimated $10K - $14K over 6-10 weeks.
See exactly what it costs to build this -- with 3 comparable funded startups.
7-day free trial. Cancel anytime.
Discover the researchers behind this paper and find similar experts.
7-day free trial. Cancel anytime.
Related Papers
- Prior WorkOnline Causal Kalman Filtering for Stable and Effective Policy OptimizationViability: 4.0
- Higher ViabilityFast and Effective On-policy Distillation from Reasoning PrefixesViability: 5.0
- Higher ViabilityLearning beyond Teacher: Generalized On-Policy Distillation with Reward ExtrapolationViability: 5.0
- Higher ViabilityEntropy-Aware On-Policy Distillation of Language ModelsViability: 7.0
- Higher ViabilityVLA-OPD: Bridging Offline SFT and Online RL for Vision-Language-Action Models via On-Policy Distillation