VISTA: Verifier-Informed Student-to-Teacher Adaptation for On-Policy Self-Distillation
On-policy self-distillation (OPSD) improves reasoning by training a problem-only student on its own rollouts using dense token-level supervision from a privileged teacher that also sees a reference solution. However, standard OPSD treats the teacher distribution as a fixed target along the student's rollout and updates only the student %, although -- even though privileged conditioning does not guarantee that the teacher always provides the most appropriate target for problem-only reasoning. This one-way supervision can therefore misdirect the student when the teacher distribution is misaligned with valid student reasoning. We therefore introduce Verifier-Informed Student-to-Teacher Adaptation (VISTA), which preserves the standard OPSD student update while using outcome-verified rollouts to adapt the teacher toward the student distribution. Within each verified rollout, VISTA further restricts this adaptation to the top-$k$ positions with the largest teacher--student KL divergence. Notably, VISTA reuses the rollout and loss function from standard OPSD, introducing no additional sampling or separate reward objective. Across AIME24, AIME25, and HMMT25 with Qwen3 models at 1.7B, 4B, and 8B, VISTA achieves the highest Avg@12 at every scale, improving over OPSD by $0.6$, $0.7$, and $2.1$ points, respectively. These results demonstrate the value of student supervision from outcome-verified rollouts and highlight student-to-teacher adaptation as a promising direction for OPSD.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
When Good Verifiers Go Bad: Self-Improving VLMs Can Regress on New Tasks
Verifier-driven self-DPO is a common recipe for self-improving production visual-language models. In this setup, a frozen verifier scores candidate generations, the top- and bottom-scoring candidates form a preference ex…
Eliciting Weak-to-Strong Generalization with On-Policy Reverse Distillation
Weak-to-strong generalization asks whether stronger models can learn from weaker supervisors and surpass them. This question is particularly important for successive model generations and multi-domain consolidation, wher…
Reward-Gated On-Policy Distillation
On-policy distillation is a powerful way to transfer reasoning ability from a strong teacher to a smaller student: the student samples trajectories from its own policy, and the teacher provides dense token-level supervis…
Persistent Teacher Anchoring for Tool-Using Agents
Distillation is common in LLM post-training, where on-policy knowledge distillation (OPKD) uses student-generated trajectories to prepare the student for downstream RL. At each state, the student matches a next-token dis…
Knowledge DistillationSG-OPD: Sign-Gated On-Policy Distillation via Sign-Consistency Gating and Phased Teacher Sampling
On-policy distillation (OPD) trains a student on its own trajectories with dense per-token supervision from a stronger teacher, and often outperforms off-policy distillation and standard reinforcement learning. However, …
Reinforcement LearningMathematical Reasoning