跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.25613 · v1

Can an LLM Fix a Language Model Without Training It?

首次提交 2026-11-19 14:59
版本更新 2026-11-19 14:59
雷达观测 2026-11-20 19:58
arXiv ↗PDF ↗

摘要

Although an external language model can select and test repairs to another model's weights, behavior can also be recovered through fixed numerical edits. We assessed the value of interactive selection in a controlled study of Transformer repair. After inspecting predictions, activation summaries, and weight statistics, the editor selects projection recipes to apply to the target without gradient updates. We analyzed 20 recorded sessions involving five masked and causal language models, comparing the selected edits with fixed projections, scripted and matched-budget black-box search, and low-rank adaptation. Under the core two-layer fault, live selection matches black-box search on SmolLM2 and Qwen2.5. For Pythia, live selection achieves 65.62% completion accuracy, compared with 0% for black-box search and 66.15% for a fixed projection. Under a separate compound fault affecting attention and feed-forward weights, both scripted and LLM-guided selection increase the descriptive five-checkpoint mean from 24.89% to 71.25% using the same operator family. On unmodified checkpoints, scripted projection yields small gains or losses on a separate set of framed questions. The experiments identify cases in which feedback changes an edit decision, operator coverage accounts for recovery, or supervised adaptation performs better. Combined with measurements of parameter restoration and preservation, these comparisons establish the scope of controller-guided repair for the tested faults and completion tasks.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-19 14:59 · 最新观测版本

Can an LLM Fix a Language Model Without Training It?

Although an external language model can select and test repairs to another model's weights, behavior can also be recovered through fixed numerical edits. We assessed the value of interactive selection in a controlled study of Transformer repair. After inspecting predictions, activation summaries, and weight statistics, the editor selects projection recipes to apply to the target without gradient updates. We analyzed 20 recorded sessions involving five masked and causal language models, comparing the selected edits with fixed projections, scripted and matched-budget black-box search, and low-rank adaptation. Under the core two-layer fault, live selection matches black-box search on SmolLM2 and Qwen2.5. For Pythia, live selection achieves 65.62% completion accuracy, compared with 0% for black-box search and 66.15% for a fixed projection. Under a separate compound fault affecting attention and feed-forward weights, both scripted and LLM-guided selection increase the descriptive five-checkpoint mean from 24.89% to 71.25% using the same operator family. On unmodified checkpoints, scripted projection yields small gains or losses on a separate set of framed questions. The experiments identify cases in which feedback changes an edit decision, operator coverage accounts for recovery, or supervised adaptation performs better. Combined with measurements of parameter restoration and preservation, these comparisons establish the scope of controller-guided repair for the tested faults and completion tasks.

查看此版本解读 →查看此版本原文 ↗