# Better Starts, Better Ends: Bootstrapped Iterative Self-Reasoning Distillation for Compressed Reasoning

arXiv: https://arxiv.org/abs/2607.15736 · HTML: https://mlsanity.com/paper/2607.15736/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning.
- Builds on 35 of its 38 references that are in the map.
- First posted on arXiv 17 Jul 2026.

## Placed in

- LLM reasoning / [Prompting and chain-of-thought reasoning](https://mlsanity.com/lr/prompting-and-chain-of-thought-reasoning/)
- Knowledge distillation / [On-policy and self-distillation for LLMs](https://mlsanity.com/distill/on-policy-and-self-distillation-for-llms/)

## Builds on (most cited first, 30 of 35)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes](https://mlsanity.com/paper/2306.13649/) (2023)
- [Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models](https://mlsanity.com/paper/2503.16419/) (2025)
- [Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models](https://mlsanity.com/paper/2601.18734/) (2026)
- [L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning](https://mlsanity.com/paper/2503.04697/) (2025)
- [TokenSkip: Controllable Chain-of-Thought Compression in LLMs](https://mlsanity.com/paper/2502.12067/) (2025)
- [Dynamic Early Exit in Reasoning Models](https://mlsanity.com/paper/2504.15895/) (2025)
- [CoT-Valve: Length-Compressible Chain-of-Thought Tuning](https://mlsanity.com/paper/2502.09601/) (2025)
- [TIP: Token Importance in On-Policy Distillation](https://mlsanity.com/paper/2604.14084/) (2026)
- [The Benefits of a Concise Chain of Thought on Problem-Solving in Large Language Models](https://mlsanity.com/paper/2401.05618/) (2024)
- [CRISP: Compressed Reasoning via Iterative Self-Policy Distillation](https://mlsanity.com/paper/2603.05433/) (2026)
- [Making Slow Thinking Faster: Compressing LLM Chain-of-Thought via Step Entropy](https://mlsanity.com/paper/2508.03346/) (2025)
- [Filter, Then Reweight: Rethinking Optimization Granularity in On-Policy Distillation](https://mlsanity.com/paper/2606.02684/) (2026)
- [OPSDL: On-Policy Self-Distillation for Long-Context Language Models](https://mlsanity.com/paper/2604.17535/) (2026)
- [Over-Reasoning and Redundant Calculation of Large Language Models](https://mlsanity.com/paper/2401.11467/) (2024)
- [Chain-of-Thought Compression Should Not Be Blind: V-Skip for Efficient Multimodal Reasoning via Dual-Path Anchoring](https://mlsanity.com/paper/2601.13879/) (2026)
- [PointCoT: A Multi-modal Benchmark for Explicit 3D Geometric Reasoning](https://mlsanity.com/paper/2602.23945/) (2026)
- [Revisiting Overthinking in Long Chain-of-Thought from the Perspective of Self-Doubt](https://mlsanity.com/paper/2505.23480/) (2025)
- [ROSD: Reflective On-Policy Self-Distillation for Language Model Reasoning across Domains](https://mlsanity.com/paper/2605.28014/) (2026)
- [Leash: Adaptive Length Penalty and Reward Shaping for Efficient Large Reasoning Model](https://mlsanity.com/paper/2512.21540/) (2025)
- [Not All Queries Need Deep Thought: CoFiCot for Adaptive Coarse-to-fine Stateful Refinement](https://mlsanity.com/paper/2603.08251/) (2026)
- [Not All Errors Are Created Equal: ASCoT Addresses Late-Stage Fragility in Efficient LLM Reasoning](https://mlsanity.com/paper/2508.05282/) (2025)
- [Do LLMs Really Need 10+ Thoughts for "Find the Time 1000 Days Later"? Towards Structural Understanding of LLM Overthinking](https://mlsanity.com/paper/2510.07880/) (2025)
- [PointRFT: Explicit Reinforcement Fine-tuning for Point Cloud Few-shot Learning](https://mlsanity.com/paper/2603.23957/) (2026)
- [Mitigating Hallucinations via Inter-Layer Consistency Aggregation in Large Vision-Language Models](https://mlsanity.com/paper/2505.12343/) (2025)
- [CtrlCoT: Dual-Granularity Chain-of-Thought Compression for Controllable Reasoning](https://mlsanity.com/paper/2601.20467/) (2026)
- [Mitigating Overthinking in Large Reasoning Models via Difficulty-aware Reinforcement Learning](https://mlsanity.com/paper/2601.21418/) (2026)

## Built on by (most cited first, 2 of 2)

- [Privileged Context as Drift in On-Policy Self-Distillation](https://mlsanity.com/paper/2610.07842/) (2026)
- [Collaborative Learning and Reasoning over Model Slices](https://mlsanity.com/paper/2610.12927/) (2026)

## Abstract (arXiv)

Large reasoning models often solve problems through long chain-of-thought (CoT) traces, yet much of this computation is spent on redundant derivations, repeated self-verification, and detours that do not improve the final answer. Existing on-policy self-distillation methods reduce this cost by matching a student model to a concise copy of itself on prefixes sampled from the student's own rollouts. We show that this objective has an initialization bottleneck. Since supervision is applied only to visited prefixes, training from a verbose base model places the KL loss on contexts that are often noisy, redundant, or already off track. In such regions, a concise teacher can provide only local corrections, while the student continues to explore trajectories that an efficient reasoner should avoid. In this paper, we propose BIRD(Bootstrapped Iterative Self-Reasoning Distillation), a two-stage self-reasoning distillation method that improves the rollout distribution before on-policy training. BIRD first samples concise solutions from the base model under a brevity instruction, keeps only answer-correct traces, and performs a lightweight prompt-switch SFT step. The traces are generated with the brevity instruction but learned under the original task prompt, turning instruction-induced conciseness into a default reasoning behavior. Starting from this warm model, BIRD then applies on-policy reverse-KL distillation with a concise self-teacher, now on cleaner and more informative prefixes. Across Qwen3 series models, BIRD achieves a stronger accuracy-efficiency trade-off than prompting and cold-start on-policy distillation on MATH-500 and AIME benchmarks. On Qwen3-8B, it improves MATH-500 accuracy from 86.2% to 92.0% while reducing the average response length from 3,099 to 1,115 tokens. These results highlight prefix support as a central factor in efficient reasoning distillation.