CircuChain: Disentangling Competence and Compliance in LLM Circuit Analysis
Mayank Ravishankara · 2026
Read CircuChain: Disentangling Competence and Compliance in LLM Circuit Analysis by Mayank Ravishankara (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-2602-15037.txt.
Description
Ravishankara introduces CircuChain, a benchmark that separates following instructions from doing physics in circuit analysis. A numerically right answer still fails if it breaks rules such as mesh direction or polarity assignments, mistakes that can spread in safety-critical systems. It is unclear whether frontier models reason from first principles or lean on training priors that clash with explicit orders. The benchmark uses paired control and trap problems across five standard layouts, varying sign rules, current directions, and polarity rules.
A multi-stage pipeline combining symbolic solvers, SPICE simulation, and an error taxonomy pins each failure to rules, physics, arithmetic, or hallucination. Across 100 tasks per model, the strongest model reasons nearly perfectly yet breaks rules often under trap conditions, while weaker models follow orders better despite weaker physics. Capability, the paper warns, does not guarantee constraint alignment.
Catalog record
| CREATOR | Mayank Ravishankara |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-15037