| title | Compress-Distill: Reasoning Trace Compression for Efficient Knowledge Distillation |
|---|---|
| authors | Maxime Griot, Paul Steven Scotti, Tanishq Mathew Abraham |
| date | 2026-06-04 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2606.05988 |
| affiliation | Université Catholique de Louvain |
| upvotes | 3 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Reasoning models produce long chain-of-thought traces that are costly to distill and encourage verbose student outputs. We study post-hoc compression of such traces before knowledge distillation. Two teachers, Qwen3.5-397B-A17B and gpt-oss-120B, generate about 283k correct traces each; two instruction-tuned models then compress them to 8.6-21.0% of their original character length. Across a 48-run main grid plus seven Qwen-teacher truncation ablations, compressed traces reduce training tokens to 12-30% of raw, speed up training by 2.0-7.6x, and shorten inference outputs by 3-19x with smaller reductions under the shorter gpt-oss teacher. However, raw traces retain the highest downstream accuracy at every scale and for both teachers. A length-matched raw-trace truncation ablation shows that compression is not merely benefiting from a smaller token budget: model-compressed traces usually beat or match naive truncation, especially for smaller students, while maintaining shorter inference outputs. Overall, reasoning-trace compression offers an accuracy-efficiency trade-off rather than a free improvement: students retain up to 96% of raw-trace accuracy while gaining up to 18x higher per-token efficiency, and at the 0.8B scale under LoRA compressed traces narrow the raw-vs-compressed gap but do not exceed raw. |
cite
@misc{griot2026compress,
title={Compress-Distill: Reasoning Trace Compression for Efficient Knowledge Distillation},
author={Maxime Griot and Paul Steven Scotti and Tanishq Mathew Abraham},
year={2026},
eprint={2606.05988},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2606.05988}
}