EGRSD and CL-EGRSD advance the accuracy-length frontier in LLM reasoning by entropy-guided weighting of token-level distillation signals from the teacher.
hub Canonical reference
CRISP: Compressed Reasoning via Iterative Self-Policy Distillation
Canonical reference. 88% of citing Pith papers cite this work as background.
20
Pith papers citing it
Background
88% of classified citations
abstract
Reasoning models think out loud, but much of what they say is noise. We introduce CRISP (Compressed Reasoning via Iterative Self-Policy Distillation), a method that teaches models to reason more concisely by distilling their own concise behavior back into themselves. The entire approach reduces to one idea: condition the same model on a ''be concise'' instruction to obtain teacher logits, and minimize per-token reverse KL on the student's own rollouts. No ground-truth answers, no token budgets, no difficulty estimators. Just self-distillation. Yet this simplicity belies surprising sophistication: CRISP automatically compresses easy problems aggressively while preserving the deliberation needed for hard ones. On Qwen3-8B and Qwen3-14B, we achieve 57--59% token reduction on MATH-500 while improving accuracy by 9--16 points absolute. On AIME 2024, the 14B model gains 10 points with 41% compression. Ablations show that qualitative conciseness instructions outperform explicit token targets, and periodic teacher refreshes yield a broad stable regime. The method generalizes across model families -- DeepSeek-R1-Distill-Llama-8B improves accuracy by up to 5 points with 17--32% compression -- and transfers beyond math to multi-step agentic planning (DeepPlanning), reducing token usage by 42--51% while preserving planning quality. Code is available at https://github.com/HJSang/OPSD_Reasoning_Compression.
hub tools
citation-role summary
background 8
citation-polarity summary
years
2026 20roles
background 8representative citing papers
Self-distillation token rewards measure input-response-feedback pointwise mutual information, and CREDIT extracts the input-specific component with contrastive baselines to improve LLM reasoning 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.
RLSD mixes self-distillation for token-level policy difference magnitudes with RLVR for reliable update directions from response correctness to reach higher convergence and better training stability.
PACED applies student pass-rate weighting w(p)=p(1-p) to distillation, concentrating on the zone of proximal development and delivering up to +8.2 gains on AIME tasks with reduced forgetting.
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.
On-policy self-distillation with teacher flip rate yields better safety-reasoning tradeoffs than off-policy or external-teacher baselines across model scales.
MSD enables cross-lingual safety transfer in LLMs via self-distillation with Dual-Perspective Safety Weighting, improving safety in low-resource languages without target response data.
A two-axis taxonomy of student entropy and teacher-student divergence identifies informative tokens in on-policy distillation, allowing near-full performance with 10-50% of tokens.
f-OPD decomposes on-policy distillation drift into rollout and supervision components, then applies a sample-level freshness score to adaptively limit stale data influence and stabilize long-horizon agent training.
Sparse rewards on capable teachers for exploration followed by dense distillation to students outperforms direct sparse reward application like GRPO on the deployment model.
Mixed-Policy Distillation transfers concise reasoning behavior from larger to smaller LLMs by having the teacher compress student-generated trajectories, cutting token usage up to 27% while raising benchmark scores.
LLM post-training is unified as off-policy or on-policy interventions that expand support for useful behaviors, reshape policies within reachable states, or consolidate behavior across training stages.
TS-OPSD internalizes temperature via on-policy self-distillation to reheat entropy-collapsed RL policies in LLMs, providing stronger initialization for further training than continued RL or rollout temperature adjustment.
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
-
Respecting Self-Uncertainty in On-Policy Self-Distillation for Efficient LLM Reasoning
EGRSD and CL-EGRSD advance the accuracy-length frontier in LLM reasoning by entropy-guided weighting of token-level distillation signals from the teacher.
-
From Generic Correlation to Input-Specific Credit in On-Policy Self Distillation
Self-distillation token rewards measure input-response-feedback pointwise mutual information, and CREDIT extracts the input-specific component with contrastive baselines to improve LLM reasoning performance.
-
Preference-Based Self-Distillation: Beyond KL Matching via Reward Regularization
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.
-
Self-Distilled RLVR
RLSD mixes self-distillation for token-level policy difference magnitudes with RLVR for reliable update directions from response correctness to reach higher convergence and better training stability.
-
PACED: Distillation and On-Policy Self-Distillation at the Frontier of Student Competence
PACED applies student pass-rate weighting w(p)=p(1-p) to distillation, concentrating on the zone of proximal development and delivering up to +8.2 gains on AIME tasks with reduced forgetting.
-
Tailoring Teaching to Aptitude: Direction-Adaptive Self-Distillation for LLM Reasoning
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.
-
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation
On-policy self-distillation with teacher flip rate yields better safety-reasoning tradeoffs than off-policy or external-teacher baselines across model scales.
-
Multilingual Safety Alignment via Self-Distillation
MSD enables cross-lingual safety transfer in LLMs via self-distillation with Dual-Perspective Safety Weighting, improving safety in low-resource languages without target response data.
-
TIP: Token Importance in On-Policy Distillation
A two-axis taxonomy of student entropy and teacher-student divergence identifies informative tokens in on-policy distillation, allowing near-full performance with 10-50% of tokens.
-
$\boldsymbol{f}$-OPD: Stabilizing Long-Horizon On-Policy Distillation with Freshness-Aware Control
f-OPD decomposes on-policy distillation drift into rollout and supervision components, then applies a sample-level freshness score to adaptively limit stale data influence and stabilize long-horizon agent training.
-
Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training
Sparse rewards on capable teachers for exploration followed by dense distillation to students outperforms direct sparse reward application like GRPO on the deployment model.
-
Reasoning Compression with Mixed-Policy Distillation
Mixed-Policy Distillation transfers concise reasoning behavior from larger to smaller LLMs by having the teacher compress student-generated trajectories, cutting token usage up to 27% while raising benchmark scores.
-
Large Language Model Post-Training: A Unified View of Off-Policy and On-Policy Learning
LLM post-training is unified as off-policy or on-policy interventions that expand support for useful behaviors, reshape policies within reachable states, or consolidate behavior across training stages.
-
Internalize the Temperature: On-Policy Self-Distillation as Policy Reheater for Reinforcement Learning
TS-OPSD internalizes temperature via on-policy self-distillation to reheat entropy-collapsed RL policies in LLMs, providing stronger initialization for further training than continued RL or rollout temperature adjustment.
-
A Brief Overview: On-Policy Self-Distillation In Large Language Models
This overview paper explains the conceptual foundations and design principles of On-Policy Self-Distillation for large language models from a beginner's perspective.
- Prefix Teach, Suffix Fade: Local Teachability Collapse in Strong-to-Weak On-Policy Distillation
- Multi-Rollout On-Policy Distillation via Peer Successes and Failures
- Adaptive Teacher Exposure for Self-Distillation in LLM Reasoning
- D-OPSD: On-Policy Self-Distillation for Continuously Tuning Step-Distilled Diffusion Models
- $\pi$-Play: Multi-Agent Self-Play via Privileged Self-Distillation without External Data