arXiv：cs.AI· Heng-Sheng Chang·· 10 小时前
Machine Deception Detector：兼顾检测准确率与不确定性的机器欺骗检测器
Accuracy and Uncertainty in Machine Deception Detection (Research White Paper)
AI 导读
该白皮书提出 Machine Deception Detector，读取语言模型全部层，仅用响应标签学习哪些 token 携带欺骗证据，并校准告警阈值使假阳性率低于预设水平。
正文
Abstract:A deception detector is only as useful as the trust its alarms deserve. Like a polygraph, detectors that read a language model's internal activations can catch deceptive responses that look honest. Yet current detectors depend on which layers they read, weaken on broader tests, and do not state how certain their alarms are. This white paper proposes the Machine Deception Detector, which targets both detection accuracy and the uncertainty of detector scores. The detector reads every layer of the language model, learns which tokens carry evidence of deception using only response labels, and calibrates its alarm threshold so that its false--positive rate stays below a prescribed level. An ensemble of such detectors measures the uncertainty of each score and raises an alarm only when its members agree, with a confidence level that sets how certain each alarm must be. The proposed evaluation trains on instructed honest and deceptive statements, tests on five unseen deceptive tasks of Liars' Bench read by six Qwen3 models from 0.6B to 32B parameters, and is designed to also show which layers carry the deception feature and how this changes with model size.
| Comments: | 18 pages, 2 figures |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.15314 [cs.LG] |
| (or arXiv:2610.15314v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.15314 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Heng-Sheng Chang [view email]
[v1]
Sun, 11 Oct 2026 03:38:54 UTC (8,947 KB)
来源：arXiv：cs.AI · arxiv.org