Pratinav Seth

← Publications / Safety Post-Training & Alignment

Cited but Not Consulted: A Counterfactual Audit of Legal Chain-of-Thought Faithfulness

Saisab Sadhu, Shreeyans Arora, Pratinav Seth

EMNLP 2026 NLLP Workshop; NeurIPS 2026 Workshop2026My role: Senior authorWorkshop
PDFarXiv
Architecture of the counterfactual audit: case facts, authority-swap prompts, internal commitment tracking with logit lens, and behavioral and mechanistic faithfulness metrics

Abstract

Large language models increasingly justify legal decisions by naming the statute or precedent behind a verdict, treated as evidence that the decision follows from it. We test this directly: holding case facts fixed, we substitute the named legal authority for an unrelated one and decode a model's evolving verdict from its hidden states. Across seven open-weight models (8B-70B) and four benchmarks spanning judicial and contractual reasoning, when explicitly required to justify a verdict by naming the governing authority, models name the correct one in 66.7%-100% of generations, while the verdict changing when the authority changes is far less consistent: 0.0%-21.7% on CaseHOLD, 30.0%-76.7% on ECHR and SCOTUS, and 43.3%-50.0% on ContractNLI. Neither scale nor a purpose-built legal-reasoning model (a best-effort LoRA reproduction; Section 6) closes this gap. A red-teaming evaluation on five core models finds compliance with an adversarial instruction hidden in the case facts (73.3%-96.4%) exceeds verdict-swap sensitivity by a wide margin, holding without exception across model rankings. Naming a legal authority is thus a poor proxy for a verdict's dependence on it, while the same verdict remains separately vulnerable to adversarial manipulation. Both findings replicate across checks ruling out prompt-wording noise and confounded sampling, and bear directly on the use of generated legal explanations as compliance or audit artefacts.

Also on OpenReview.

Cite

@inproceedings{sadhu2026cited,
  title         = {{Cited but Not Consulted: A Counterfactual Audit of Legal Chain-of-Thought Faithfulness}},
  author        = {Saisab Sadhu and Shreeyans Arora and Pratinav Seth},
  booktitle     = {EMNLP 2026 NLLP Workshop; NeurIPS 2026 Workshop},
  year          = {2026},
  eprint        = {2610.12361},
  archivePrefix = {arXiv},
  url           = {https://arxiv.org/abs/2610.12361}
}
All →