Cool paper on catching alignment failures with Jev. The ideas is to ask Jev one generic yes/no question about a model's response, and use it…

Cool paper on catching alignment failures with Jev. The ideas is to ask Jev one generic yes/no question about a model's response, and use its probability as a score. With no extra training, that score separates failures from good responses well, with a median AUROC of 0.886. It is also cheap. On 19 benchmarks, a Jev pass cost 0.30, while the LLM judges those benchmarks use cost 18.96. Researchers built RLCDAlignBench from 44 existing benchmarks across ten failure types, including sycophancy, jailbreaks, deception, prompt injection and reward hacking. Jev, TypeSafe AI's calibrated decision model, answers many typed questions about one input in a single call, each with a probability. On StrongREJECT, it agrees with human labels as well as the GPT-4o-mini scorer does, and ranks responses better (AUROC 0.971 vs 0.929). Where Jev confidently disagreed with benchmark labels, it found label errors in three benchmarks. They find that the question wording matters little. The thresholds do not transfer between benchmarks, and fitting one on 10 labelled items raises median F1 from 0.706 to 0.793. Paper: https://arxiv.org/abs/2609.29429 Chat with Paper: https://academy.dair.ai/papers/just-ask-jev-reinforcement-learning-for-calibrated-decisions-as-a-zero-shot-dete-2609.29429

Related

• // Evaluating and Improving LLM Self-Modeling // Really interesting paper. Can a model answer questions about its own behavior? The question…
• Cool to see classification models resurge in the era of agents. The effects of Jev are insane. Also very cool to see our emotion dataset bui…

Source: DAIR.AI (X) | 2026-09-27