Quantifying Cross-Query Contradictions in Multi-Query LLM Reasoning
For developers of LLM-based systems that handle multiple interdependent queries, this work provides a method to enforce global logical consistency, which is critical for applications like legal reasoning or question answering.
The paper addresses the problem of cross-query contradictions in LLM reasoning, introducing a benchmark and metrics to measure global consistency. Their solver-augmented method improves SetCons from 0.56 to 0.94 across four domains while maintaining per-query accuracy.
Large language models frequently produce mutually inconsistent answers when reasoning over multiple related queries. We study case-file logical consistency: maintaining a globally satisfiable belief state across interdependent queries. We introduce a benchmark of 390 multi-query reasoning instances with entailment/contradiction/unknown labels and propose set-level metrics including Case Satisfiability Rate, Contradiction Density and Revision Cost. Our solver-augmented approach extracts commitments, verifies global satisfiability and performs counterexample-guided repair. Across four reasoning domains, our method substantially reduces cross-query contradictions (SetCons: 0.56 to 0.94) while preserving per-query accuracy, demonstrating that global coherence is critical for robust multi-query reasoning.