QuoteBench asks a deployment question that matched scores can hide: did the model generate a bad Bash command, or did the execution interface break it afterward? Across 56 execution-verified tasks, replaying the same reply through one added parser lowers success by 55.4–73.2 percentage points. Disclosing that boundary recovers 30.4–60.7 points in six of eight configurations. The results show why coding-agent evaluations should report the generation contract, execution path, operating point, and final-state validator—not just a matched score.</p>\n","updatedAt":"2026-08-21T15:06:32.031Z","author":{"_id":"6745c27763ec249909b1d9af","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6745c27763ec249909b1d9af/fw0A1HUPh7a-wKGmdKgLd.png","fullname":"lsamc","name":"lsamc","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9194616079330444},"editors":["lsamc"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6745c27763ec249909b1d9af/fw0A1HUPh7a-wKGmdKgLd.png"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.13547","authors":[{"_id":"6a80a40bb601d59c65281166","user":{"_id":"6745c27763ec249909b1d9af","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6745c27763ec249909b1d9af/fw0A1HUPh7a-wKGmdKgLd.png","isPro":false,"fullname":"lsamc","user":"lsamc","type":"user","name":"lsamc"},"name":"Shangao Li","status":"claimed_verified","statusLastChangedAt":"2026-08-21T13:17:13.660Z","hidden":false},{"_id":"6a80a40bb601d59c65281167","name":"Yao Zhang","hidden":false},{"_id":"6a80a40bb601d59c65281168","name":"Volker Tresp","hidden":false},{"_id":"6a80a40bb601d59c65281169","name":"Yuanyuan Yang","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/6745c27763ec249909b1d9af/nOzdb9CM5P7gfr2GhJmrP.png","https://cdn-uploads.huggingface.co/production/uploads/6745c27763ec249909b1d9af/6Fyh5Oh-mgzO1HHdF3f9I.png"],"publishedAt":"2026-08-13T00:00:00.000Z","submittedOnDailyAt":"2026-08-21T00:00:00.000Z","title":"QuoteBench: How Matched Scores Can Hide Command-Path Failures","submittedOnDailyBy":{"_id":"6745c27763ec249909b1d9af","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6745c27763ec249909b1d9af/fw0A1HUPh7a-wKGmdKgLd.png","isPro":false,"fullname":"lsamc","user":"lsamc","type":"user","name":"lsamc"},"summary":"LLM coding agents issue Bash commands through interfaces that may serialize, wrap, and reparse model output. Matched execution scores alone cannot distinguish command-generation errors from failures introduced after generation. QuoteBench measures this boundary with exact final-state validation on 56 one-shot tasks from 14 incident-derived families, crossing the generation contract with the execution transport around one deliberately unescaped added parser. Escaping at the interpolation point reproduces each replayed reply's raw-path outcome, so any recovery under a disclosed boundary must come from the model changing its generation. Across eight same-window configurations, replaying the same reply through the added parser lowers success by 55.4 to 73.2 percentage points; disclosure recovers 30.4 to 60.7 points for six configurations, and zero or slightly negative for the other two. Raw generation is nearly saturated at the frontier; boundary adaptation is what still separates models. GPT-5.6-sol's matched gap of -3.6 points hides -64.3 points of damage and +60.7 points of compensation. The deployment configuration reorders models: one reversal among 26 comparable pairs is unambiguous and four more sit on single-task margins. Evaluations of command-issuing agents should report the model configuration, generation contract, execution path, operating point, and final-state validator rather than treat a matched score as an intrinsic model property.","upvotes":2,"discussionId":"6a80a40bb601d59c6528116a","projectPage":"https://quotebench.lsamc.website/","githubRepo":"https://github.com/LeonardNJU/quoteBench","githubRepoAddedBy":"user","ai_summary":"QuoteBench reveals that execution-boundary parsing errors significantly reduce LLM coding agent success, and disclosing the boundary helps recover performance, showing that evaluation must account for deployment configuration rather than treating matched scores as intrinsic model properties.","ai_keywords":["LLM coding agents","Bash commands","QuoteBench","generation contract","execution transport","parser","boundary adaptation","final-state validation"],"ai_summary_model":"thinkingmachines/Inkling-Small","githubStars":3},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6745c27763ec249909b1d9af","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6745c27763ec249909b1d9af/fw0A1HUPh7a-wKGmdKgLd.png","isPro":false,"fullname":"lsamc","user":"lsamc","type":"user"},{"_id":"63ac5701c21e60a3e9b58aa7","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63ac5701c21e60a3e9b58aa7/g6EX7diOpuA94R2ab-rZC.png","isPro":true,"fullname":"Dipankar Sarkar","user":"dipankarsarkar","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.13547.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2608.13547 in a model README.md to link it from this page.

Cite arxiv.org/abs/2608.13547 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.