A stop-gradient consistency regularizer mitigates context-induced degradation in on-policy distillation, improving robustness across 12 configurations.
hub Canonical reference
Unifying group-relative and self-distillation policy optimization via sample routing
Canonical reference. 100% of citing Pith papers cite this work as background.
21
Pith papers citing it
Background
100% of classified citations
hub tools
citation-role summary
background 6
citation-polarity summary
years
2026 21verdicts
UNVERDICTED 21roles
background 6polarities
background 6representative citing papers
OPRD performs distillation in hidden-state space on on-policy data for deterministic gradients and better math benchmark performance, plus OPRD-Bridge for cross-architecture transfer via low-rank projectors.
DistIL applies distributional DAgger with forward cross-entropy to achieve monotonic policy improvement and better Pass@N from rich feedback in RL for reasoning tasks.
EST-PRM stress-tests five PRM models on 4,687 reasoning chains from MATH-500, GSM8K, and PRMBench using three label-preserving transformations and reports model-specific vulnerability patterns.
RLRT augments GRPO by reinforcing tokens on correct student rollouts that the teacher would not have predicted, outperforming standard self-distillation and exploration baselines on Qwen3 models.
TRACE improves math reasoning by distilling only on annotator-marked critical spans with forward KL on correct key spans, optional reverse KL on errors, and GRPO elsewhere, gaining 2.76 points over GRPO while preserving OOD performance.
PBSD derives a reward-reweighted teacher distribution as the analytic optimum of a reward-regularized objective, yielding better stability and performance than KL-based self-distillation on math reasoning and tool-use tasks.
RAPS-DA improves RAG robustness to heterogeneous knowledge conflicts by training regime-specific peer specialists with hard routing and a dual-layer token selector for focused supervision.
RLCSD contrasts teacher-student distributional gaps under correct versus wrong hints to suppress privilege-induced style drift and concentrate supervision on task tokens, outperforming GRPO and prior OPSD on Qwen3 and Olmo models.
PTD-PO supplies step-wise token-distribution supervision to student policies via in-context privileged hints derived from spatial attention and intermediate reasoning, while keeping the student in an answer-free context and using Top-K Jensen-Shannon divergence for stable alignment.
SocraticPO adds Socratic-style teacher guidance and reward decay to RL rollouts for LLMs, improving performance on scientific reasoning benchmarks over baselines.
DASD improves math reasoning in LLMs by adaptively directing self-distillation based on per-token entropy to balance exploration and step accuracy, outperforming prior self-distillation and RLVR baselines on six benchmarks.
RESD turns failure trajectories into token-level supervision via retrospective reflections and a persistent global playbook, enabling faster improvement than standard self-distillation or GRPO with only one rollout per prompt.
Anti-Self-Distillation reverses self-distillation signals via PMI to fix overconfidence on structural tokens, matching GRPO baseline accuracy 2-10x faster with up to 11.5 point gains across 4B-30B models.
SOD reweights on-policy distillation strength step-by-step using divergence to stabilize tool use in small language model agents, yielding up to 20.86% gains and 26.13% on AIME 2025 for a 0.6B model.
SimCT enlarges the supervision space in cross-tokenizer on-policy distillation using short jointly tokenizable multi-token continuations, producing consistent gains over shared-token baselines on math and code benchmarks.
On-policy distillation works when student and teacher models share thinking patterns and the teacher adds new capabilities, with success tied to alignment on a small set of high-probability tokens.
DemoPSD uses a reverse-KL barycenter target modulated by distribution discrepancy for selective teacher guidance in LLM self-distillation, claiming leakage attenuation, exploration preservation, and superior performance on SciKnowEval and GPQA.
VISD proposes structured self-distillation with a multi-dimensional judge model and direction-magnitude decoupling to improve token-level credit assignment and convergence speed in VideoLLM reasoning training.
PGPO modulates per-step trust in self-distilled updates via a mutual-information estimate derived from a viscous-fluid analogy, preserves SGD weak-approximation order, and reports gains of up to 4.5 points on Science-QA while avoiding late-training collapse.
This overview paper explains the conceptual foundations and design principles of On-Policy Self-Distillation for large language models from a beginner's perspective.
citing papers explorer
-
VISD: Enhancing Video Reasoning via Structured Self-Distillation
VISD proposes structured self-distillation with a multi-dimensional judge model and direction-magnitude decoupling to improve token-level credit assignment and convergence speed in VideoLLM reasoning training.