Distill Skills into Weights, Not Prompts: Abstract Skills as Privileged Signals for On-Policy Self-Distillation
Yubo Jiang, Fengying Xie, Zhiguo Jiang, Haopeng Zhang
Abstract
Reinforcement learning with verifiable rewards yields no group-relative signal when rollout groups are uniformly correct or uniformly wrong, which account for 63.0-68.0% of groups in our experiments. We propose SKALD (Skill-Anchored Latent Distillation), an on-policy self-distillation framework that uses two context views of the same Qwen3-Base model: a question-only student and a teacher conditioned on an abstract, explicit-answer-filtered skill card. The student is trained on its own prefixes, transferring the skill-induced advantage into shared parameters without privileged input at test time. To stabilize context-induced distribution mismatch, SKALD employs an annealed exponentially tilted objective that downweights teacher-preferred tokens with very low student likelihood; as the tilt vanishes, it converges to teacher cross-entropy and recovers the forward-KL student gradient. An empirical gate activates distillation only when verified rollouts estimate a positive teacher advantage. Across five held-out mathematics benchmarks, SKALD improves overall avg@8 over GRPO by +2.46, +4.85, and +12.01 at 0.6B, 1.7B, and 4B, respectively. At 1.7B, zero-variance-only distillation recovers 84.7% of the full gain, while SKALD remains +4.06 above FLOP-matched GRPO and exceeds contextual skill exposure by +3.77. These results show that abstract skills provide dense supervision where group-relative rewards become uninformative.
Create a lesson
Related papers
How Model Growth, Recursion, and Boundary Operators Influence Scaling Exponents
Zixi Chen, Akshay Vegesna, Samip Dahal et al.
Evidence-Grounded Agentic Formulation Development in an Autonomous Laboratory
Michael M. Craig, Riley J. Hickman, Yingshan Ma et al.
Probabilistic Linear Explanations
Frederic Koriche, Jean-Marie Lagniez, Chi Tran
Double descent is the principle of least action
Congzhou M Sha
RLLBC-Lib: An Educational Code Library for Reinforcement Learning and Learning-Based Control
Bernd Frauenknecht, Emma Cramer, Artur Eisele et al.
Higher-order pruning of experts in mixture-of-experts language models
Alex M. Tseng, Prannay Kaul, Luca Zancato et al.