arXiv 2609.29429 · cs.AI · 2026-09-24
Just Ask Jev: Reinforcement Learning for Calibrated Decisions as a Zero-Shot Detector of AI Alignment Failures
Ruoqi Guo, Yi Liu, Gelei Deng, Yuekang Li, Lida Zhao, Yutao Wu, Simin Chen, Ying Zhang and 1 more
Affiliation unverified
In latest · #170 Added on 2026-10-21
One-line takeaway
Tests whether Jev, a calibrated decision model, can zero-shot detect ten alignment failures such as sycophancy, jailbreaks and deception.
测试经 RL 校准的决策模型 Jev 能否零样本检测谄媚、越狱、欺骗等十类对齐失败。
Read on arXiv PDF Hugging Face discussion Code Project page
Topics & disciplines
More on Safety & interpretability
- Behavioral Compositionality of Sparse Autoencoder Representations ▲110
- DecepEval: A Benchmark for Evaluating Deception in LLM Agents ▲85
- Lexical Reference Frames: Interpreting LM Residual Streams in Native Semantics ▲55
- U-Space: Uncovering When and Why Uncertainty Arises in Language Models ▲54
- APEX-Harm: Evaluating Agent Safety in Shared Professional Workspaces ▲26
About this data
Metadata from the arXiv API (verified); upvotes from Hugging Face Daily Papers; citations from Semantic Scholar. Last refreshed 2026-10-22.
The takeaway is written by AIMap from the abstract; the abstract itself is not reproduced — read the original.