FormalJudge : A Neuro-Symbolic Paradigm for Agentic Oversight
Jiayi Zhou, Yang Sheng, Hantao Lou, Yaodong Yang, Jie Fu · 2026
Read FormalJudge : A Neuro-Symbolic Paradigm for Agentic Oversight by Jiayi Zhou, Yang Sheng, Hantao Lou, Yaodong Yang, Jie Fu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-11136.txt.
Description
Who watches the watchers when judges are themselves chance-based? Jiayi Zhou and four co-authors propose FormalJudge, a neuro-symbolic frame for agentic oversight. Language models first turn high-level intent into small, checkable constraints, then prove compliance with Dafny specs and a satisfiability solver.
That two-way Formal-of-Thought design replaces chance-based scores with math guarantees. Tests span behavioural safety, multi-domain constraint following, and deception finding. The frame beats language-model-judge rivals by a clear margin, lets a small judge catch deception from far larger agents, and gains steadily with iterative refinement. The work argues that formal checks can escape the judge's own failure modes.
Catalog record
| CREATOR | Jiayi Zhou, Yang Sheng, Hantao Lou, Yaodong Yang, Jie Fu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 73 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-11136