@rohanpaul_ai: New Meta paper finds that AI agents improve more from their past attempts when each new problem is matched to notes fro…
Summary
A new Meta paper shows AI agents learn better from past attempts when experiences are routed into per-problem-type memory clusters rather than a single shared history, boosting performance by 15-30% of the max score while cutting API costs by 65-72%.
View Cached Full Text
Cached at: 11/02/26, 10:18 PM
New Meta paper finds that AI agents improve more from their past attempts when each new problem is matched to notes from similar problems, instead of one shared record of everything they have done.
A simple way to let agents learn from past runs is one shared history. With mixed tasks, it gets costly and mixes unrelated experience.
Routed CL adds an LLM router that reuses a memory cluster or starts a new one before each run. Each cluster uses 1 of 5 memory types, such as raw past runs or distilled lessons.
Across 398 mixed problems and 4 models, the shared history barely beat no memory, with a mean gain of 0.015 out of 1. Routed CL beat it by 15% to 30% of the maximum possible score and cut estimated API cost by 65% to 72%.
If your agents handle mixed task types, split their memory into separate pools by problem type instead of sharing one history.
Similar Articles
@rohanpaul_ai: Meta paper shows that coding agents get much better when they reuse short summaries of past attempts instead of raw log…
A Meta paper shows that coding agents improve significantly when they reuse short summaries of past attempts instead of raw logs, achieving strong gains on SWE-Bench and Terminal-Bench with Claude 4.5 Opus.
@rohanpaul_ai: New Tencent paper letting AI auto-improve your agent's instructions works better and costs far less if it remembers pas…
Tencent's paper SkillAdam improves agent skill-evolution by having the rewriting AI keep a memory of past fixes and make smaller edits when results are mixed, achieving 28.3% accuracy vs. SkillOpt's 21.7% while using about a third of the tokens on long shopping and travel-planning tasks.
@rohanpaul_ai: New Meta paper shows that frontier agents differ widely in how they learn from experience, and the best performer is no…
Meta paper evaluates frontier agents on how well they learn from experience across chess, Go, Hex and NetHack with frozen weights, defining plasticity as held-out gain per dollar. Claude Fable 5 showed the largest absolute gains while GPT-5.6 Sol was far more cost-efficient, showing the best performer isn't the most efficient learner.
@rohanpaul_ai: New Meta paper. Most agent memory systems use one confidence cutoff to decide which facts about a user to save, but thi…
A new Meta paper finds that agent memory systems work better by applying a stricter confidence cutoff only to user value and belief claims, rather than using a single global threshold. A targeted bar on values reduced unsupported saved facts from 6.2% to 4.0% while retaining about 13 points more of the source's key facts.
@rohanpaul_ai: Meta dropped a great paper. AI agents perform better and cost less when they manage their own context, so stop hard-cod…
Meta released a paper on Context Language Models (CLMs), which let AI agents manage their own context rather than relying on hard-coded cleanup rules, achieving an 11.4% improvement over Codex-style summarization while using 21.5% less compute.