Metraipolis › archive › dc_7557024cefcc
Memory Augmentation Unlocks Efficient Chain-of-Thought Reasoning — arXiv:2608.21265
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.21265
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T10:18:34.319Z
- SHA-256 of the archived text
4a956beaeeb44a2824e006580bfff2318058586b0ab75a479690717c30a8f958- Stable URL of this page
https://metraipolis.com/documento/dc_7557024cefcc
How to cite
Memory Augmentation Unlocks Efficient Chain-of-Thought Reasoning — arXiv:2608.21265 — arXiv (en), https://arxiv.org/abs/2608.21265 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7557024cefcc (Metraipolis archive dc_7557024cefcc, sha256 4a956beaeeb44a2824e006580bfff2318058586b0ab75a479690717c30a8f958)
BibTeX
@misc{dc7557024cefcc,
title = {Memory Augmentation Unlocks Efficient Chain-of-Thought Reasoning — arXiv:2608.21265},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.21265},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7557024cefcc; sha256 4a956beaeeb44a2824e006580bfff2318058586b0ab75a479690717c30a8f958},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.21265. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models often rely on Chain-of-Thought (CoT) reasoning to solve complex tasks, but verbose reasoning traces introduce substantial inference overhead. CoT compression shortens generation, yet aggressive compression may disrupt logical coherence and degrade performance. We formalize this trade-off as the Context-Generation Substitution Law, where explicit reasoning context substitutes for part of decode-time generation. Based on this principle, we propose Memory-Augmented Compression, a training-free framework that constructs reusable reasoning memories from historical traces and retrieves them as prefill-side scaffolds. Rather than using raw demonstrations, these memories summarize reusable reasoning patterns, key constraints, and critical operations to compensate for information lost during compression. Experiments show that Memory consistently improves prompt-based Chain-of-Draft (CoD) compression across mathematical reasoning, complex reasoning, and science question answering tasks, yielding accuracy gains of 21.4, 28.0, 29.5, and 6.61 points over CoD on GSM8K, MATH, BBH, and MMLU-Sci, while achieving a 1.14-1.49x latency speedup latency speedup over standard CoT. Memory is also compatible with token-level, reasoning-trace-level, and inference-state compression mechanisms. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-24 in cs.CL). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.21265 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-18T10:17:37.603Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).