Yunfeng Zhu, Shengcheng Yu, Zihan Li, Zhenyu Chen · 2026 · arXiv
DOI: 10.48550/arXiv.2610.29609 · open-access PDF
Dalam pipa otomatisasi, laporan kesuksesan agen GUI yang percaya diri dapat membiarkan tahap-tahap downstream melanjutkan bahkan ketika jejak eksekusi yang tercatat tidak menunjukkan bahwa agen menghasilkan hasil yang diminta. Kami menyebut kasus ini laporan yang salah berbahaya: klaim keberhasilan yang tidak terjamin bahwa jejak tidak mendukung sebagai pencapaian kausal selama berjalan. Kami mempelajari audit laporan - jejak, yang memeriksa klaim penyelesaian agen terhadap jejak yang tercatat secara independen. Intervensi sadar / buta terkontrol mengungkapkan kebocoran laporan diri: laporan diri sendiri agen mengubah bagaimana penyelesaian dinilai.
Machine translation (Meta M2M100); the English original follows.
In automation pipelines, a GUI agent's confident success report can let downstream stages proceed even when the recorded execution trace does not show that the agent produced the requested outcome. We call this case a dangerous misreport: an unhedged success claim that the trace does not support as causal accomplishment during the run. We study report--trace auditing, which checks an agent's completion claim against an independently recorded trace. A controlled aware/blind intervention reveals self-report leakage: the agent's own self-report changes how completion is judged. The same judge rea
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.29609, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange