Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.14525 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915939558096896 |
|---|---|
| author | Salla, Rohit Kumar Amancherla, Ramya Manasa Saravanan, Manoj |
| author_facet | Salla, Rohit Kumar Amancherla, Ramya Manasa Saravanan, Manoj |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_14525 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Quantifying Cross-Query Contradictions in Multi-Query LLM Reasoning Salla, Rohit Kumar Amancherla, Ramya Manasa Saravanan, Manoj Artificial Intelligence 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. |
| title | Quantifying Cross-Query Contradictions in Multi-Query LLM Reasoning |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2604.14525 |