跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2607.12713 · v2

Code-Anchored Review Feedback Improves Correctness and Lowers Hosted Editing Cost on Large R Files

首次提交 2026-07-14 20:40
版本更新 2026-11-15 09:54
雷达观测 2026-11-17 12:06
arXiv ↗PDF ↗

摘要

Tokens drive the cost and latency of generative AI (GAI) code editing, and correctness decides whether an edit is useful. We study how the format of a reviewer's feedback, a human-AI collaboration (HAIC) channel, impacts both. In a paired experiment on 36 synthetic base-R tasks with seven models, the same requested changes were delivered as a holistic list (control) or as the export of FileMark (treatment). That code-review extension anchors each comment to the code it concerns and quotes it. For the typical task, total tokens changed by -4.2% to +14.7% across models, while output tokens fell by 14.7% (Claude Opus) and 50.5\% (Claude Sonnet). Since an output token costs five times as much as an input token, the hosted cost of an edit on files over 100 lines fell by 20.0% (Opus) and 46.3% (Sonnet). The treatment raised the correct-run share of four of five local models by 24% to 46%. A decomposition experiment places these gains in the format, not the wording, and mostly in the quoted code. Misplaced anchors cut correctness sharply. Code-anchored feedback thus lowers hosted editing cost on large R files and makes weaker models more correct, provided the anchors are accurate.

作者备注:Code, task bank, and raw model outputs: https://github.com/billyl320/filemark_usage

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

本站首次收录为 v2,暂无更早版本可对照。

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。

v2 · 2026-11-15 09:54 · 最新观测版本

Code-Anchored Review Feedback Improves Correctness and Lowers Hosted Editing Cost on Large R Files

Tokens drive the cost and latency of generative AI (GAI) code editing, and correctness decides whether an edit is useful. We study how the format of a reviewer's feedback, a human-AI collaboration (HAIC) channel, impacts both. In a paired experiment on 36 synthetic base-R tasks with seven models, the same requested changes were delivered as a holistic list (control) or as the export of FileMark (treatment). That code-review extension anchors each comment to the code it concerns and quotes it. For the typical task, total tokens changed by -4.2% to +14.7% across models, while output tokens fell by 14.7% (Claude Opus) and 50.5\% (Claude Sonnet). Since an output token costs five times as much as an input token, the hosted cost of an edit on files over 100 lines fell by 20.0% (Opus) and 46.3% (Sonnet). The treatment raised the correct-run share of four of five local models by 24% to 46%. A decomposition experiment places these gains in the format, not the wording, and mostly in the quoted code. Misplaced anchors cut correctness sharply. Code-anchored feedback thus lowers hosted editing cost on large R files and makes weaker models more correct, provided the anchors are accurate.

查看此版本解读 →查看此版本原文 ↗