@rohanpaul_ai: Nice new paper on continuous improvement of AI agents from Princeton and Meta Finds that agents repeat fewer mistakes w…

X AI KOLs Timeline Papers
ai-agents continual-learning lora fine-tuning princeton meta gaia-benchmark

Summary

Princeton 与 Meta 的新论文提出 Self-Direction 方法：让失败的 AI 智能体反思并将教训训练成约 0.02% 大小的 LoRA 适配器，由路由器在遇到相似任务时自动加载，从而显著减少重复错误。在 Qwen3.6-27B 上，GAIA 任务成功率从 38.8% 提升至 55.8%，且跨 185 个早期错误的修复持久性远超普通重试。

Nice new paper on continuous improvement of AI agents from Princeton and Meta Finds that agents repeat fewer mistakes when each fix is saved as a LoRA adapter and reused on similar tasks. Agents can fix a mistake within a session, then often repeat it in the next one. Notes in context get forgotten, and full fine-tuning is costly and can disrupt other skills. With Self-Direction, a failing agent reflects on its run and trains the lesson into a LoRA adapter about 0.02% of the model's size. A router later loads the closest adapter when the agent fails a similar task. With Qwen3.6-27B, task success on GAIA rose from 38.8% to 55.8%, beating in-context reflections and a Skills.md file. Across 185 earlier mistakes, 67.3% of repairs still worked by the third encounter, against 42.9% for plain retrying. If your agents keep making the same mistakes, save each fix as a small adapter they can reload, not only as prompt notes.
Original Article
View Cached Full Text

Cached at: 10/30/26, 09:09 AM

Nice new paper on continuous improvement of AI agents from Princeton and Meta

Finds that agents repeat fewer mistakes when each fix is saved as a LoRA adapter and reused on similar tasks.

Agents can fix a mistake within a session, then often repeat it in the next one. Notes in context get forgotten, and full fine-tuning is costly and can disrupt other skills.

With Self-Direction, a failing agent reflects on its run and trains the lesson into a LoRA adapter about 0.02% of the model’s size. A router later loads the closest adapter when the agent fails a similar task.

With Qwen3.6-27B, task success on GAIA rose from 38.8% to 55.8%, beating in-context reflections and a Skills.md file. Across 185 earlier mistakes, 67.3% of repairs still worked by the third encounter, against 42.9% for plain retrying.

If your agents keep making the same mistakes, save each fix as a small adapter they can reload, not only as prompt notes.

Similar Articles