{"slug": "ctwm-memory-controller-cuts-longmemeval-tokens-24-48-with-accuracy-parity", "title": "CTWM memory controller cuts LongMemEval tokens 24.48% with accuracy parity", "summary": "A rank-based memory controller called CTWM cut prompt tokens by 24.48% on LongMemEval with aggregate accuracy parity, according to an arXiv paper (2610.00010) on heavy-tailed memory traces in long-horizon language agents. The same controller reduced tokens by 5.9% on Synthetic Graph World while lowering bottom-half tail prediction error by 13.6%. The paper's stated takeaway is that agent memory retrieval should be audited for core–tail concentration, since semantic policies can overuse a small memory core and silently accumulate errors on rare states, and that allocating context by retrieval rank while retaining summarized tail state can cut cost without sacrificing coverage.", "body_md": "[arXiv](https://arxiv.org/abs/2610.00010)\n\n### CTWM memory controller cuts LongMemEval tokens 24.48% with accuracy parity\n\nWhich summary reads better? Pick one — models revealed after.Both summaries are AI-generated.\n\nMistral Large quota or rate limit — check usage and plan. Original headline: Heavy-Tailed Memory Traces in Long-Horizon Language Agents\n\nA rank-based memory controller cut prompt tokens by 24.48% on LongMemEval with aggregate accuracy parity, and by 5.9% on Synthetic Graph World while reducing bottom-half tail prediction error by 13.6%. The practical takeaway is that agent memory retrieval should be audited for core–tail concentration, because semantic policies can overuse a small memory core and silently accumulate errors on rare states; allocating context by retrieval rank while retaining summarized tail state can reduce cost without sacrificing coverage.", "url": "https://wpnews.pro/news/ctwm-memory-controller-cuts-longmemeval-tokens-24-48-with-accuracy-parity", "canonical_source": "https://www.snipvote.com/story/cmuqn2cct0006t56bzh30yjna", "published_at": "2026-10-02 07:37:52.071567+00:00", "updated_at": "2026-10-02 07:37:53.559323+00:00", "lang": "en", "topics": ["ai-agents", "large-language-models", "ai-research", "ai-infrastructure"], "entities": ["CTWM", "LongMemEval", "Synthetic Graph World", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ctwm-memory-controller-cuts-longmemeval-tokens-24-48-with-accuracy-parity", "markdown": "https://wpnews.pro/news/ctwm-memory-controller-cuts-longmemeval-tokens-24-48-with-accuracy-parity.md", "text": "https://wpnews.pro/news/ctwm-memory-controller-cuts-longmemeval-tokens-24-48-with-accuracy-parity.txt", "jsonld": "https://wpnews.pro/news/ctwm-memory-controller-cuts-longmemeval-tokens-24-48-with-accuracy-parity.jsonld"}}