# 自我改进智能体记忆奖励膨胀研究

- 来源：Rohan Paul (@rohanpaul_ai)
- 发布时间：2026-08-09 20:52
- AIHOT 分数：47
- AIHOT 链接：https://aihot.virxact.com/items/cmsltqsdj030wrovir9rcabai
- 原文链接：https://x.com/rohanpaul_ai/status/2086435490119713248

## AI 摘要

一项研究发现，自我改进的LLM智能体即使冻结权重，也会因信任自身记忆而性能退化。智能体存储过往片段并用LLM评分复用，但测试模型对其错误答案的自评分中，有31%至54%被误判为正确。作者将这一现象称为“回声差距”，且更强或不同的LLM无法可靠修复此问题。

## 正文

A self-improving agent can keep its weights frozen and still get worse through the memories it learns to trust.

These agents store past episodes, score them with an LLM, and reuse them later as precedents without changing model weights.

Wrong episodes can still receive high self-scores; across the paper's factual banks, the tested models endorsed 31% to 54% of their own wrong answers as correct.

Once that score enters persistent memory, the mistake can influence future decisions instead of disappearing after one bad answer.

The authors call this the Echo Gap.

Stronger or different LLMs did not reliably repair it, because their grading errors often remained correlated with the original self-grading bias.

- arxiv. org/abs/2608.00017

Title: "Memory Reward Inflation in Self-Improving LLM Agents"