← 研踪 PaperTrailarXiv:2610.36717 · v1
A Harness Reducing over-Defensive Writing in Agents by Isolating Objections
首次提交 2026-10-27 14:32
版本更新 2026-10-27 14:32
雷达观测 2026-10-29 00:35
摘要
When expanding a report, writing with agents may introduce a stronger claim than the article makes and then disclaim it. This is called over-defensive writing. We assume that this is because the agent is writing for responding to an adversarial AI reviewer, instead of a human reader. The defensive context may also influnce the consequent when we simply ban defensive writing in the output through prompts. In order to reduce over-defensive writing, we produce a harness that separates the reviewer, filter, and writer. The writer agent receives only the problems choosen by the filter, which is an independent context. We evaluate this harness on 25 business-project drafts expanded into formal reports, the result shows that the defensive propositions decreased from 59 to 2, which is a 96.61\% reduction, no unsupported factual additions were identified while the source-claim retention rises from 80.35% to 93.12%.
作者备注：defensive writing; context isolation; harness
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 · 2026-10-27 14:32 · 最新观测版本
A Harness Reducing over-Defensive Writing in Agents by Isolating Objections
When expanding a report, writing with agents may introduce a stronger claim than the article makes and then disclaim it. This is called over-defensive writing. We assume that this is because the agent is writing for responding to an adversarial AI reviewer, instead of a human reader. The defensive context may also influnce the consequent when we simply ban defensive writing in the output through prompts. In order to reduce over-defensive writing, we produce a harness that separates the reviewer, filter, and writer. The writer agent receives only the problems choosen by the filter, which is an independent context. We evaluate this harness on 25 business-project drafts expanded into formal reports, the result shows that the defensive propositions decreased from 59 to 2, which is a 96.61\% reduction, no unsupported factual additions were identified while the source-claim retention rises from 80.35% to 93.12%.
查看此版本解读 →查看此版本原文 ↗