Papers tracker

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.

Large language models are increasingly used in legal research and drafting, but they can still produce claims that sound convincing without being supported by the cited source. We introduce PARCEL, a benchmark for checking whether a legal claim is supported by the underlying authority. Using recent New York State Court of Appeals decisions, we build a dataset of 3,396 parenthetical-style claims labeled as Supported, Refuted, or Not Found. We cast this task as a three-way natural language inference problem and evaluate several state-of-the-art LLMs in a zero-shot setting. Although the strongest models reach up to 0.97 accuracy, the results also show an important weakness: models still incorrectly mark unsupported claims as supported, even when the full opinion text is provided. Across models, missing support is harder to detect than direct contradiction, and fabricated but plausible citations cause the largest drop in performance. Overall, PARCEL provides a practical benchmark for testing claim-level groundedness in legal RAG systems.

On-chain transfer volume is often used to describe activity in tokenized real-world assets (RWAs), although recorded transfers may reflect very different economic functions, including token creation, destruction, issuer-related operations, and movements between other addresses. Ownership measures face a similar problem because a large address balance may belong to a pool or intermediary that represents multiple investors. We address these two measurement problems by decomposing transfer activity and making ownership coverage explicit. Transfer turnover is calculated before and after observable event categories are separated, using the same asset, observation window, and supply denominator. For ownership, we compare concentration measured against total supply with concentration among the addresses that remain after identified intermediaries are excluded. Using Ethereum-based tokenized funds from December 2025 to August 2026, we find that transfer classification can materially change measured turnover. We also show that removing an intermediary can lower concentration when measured against total supply while concentration among the remaining addresses increases. Transfer activity and address concentration need to be interpreted with their underlying transaction functions and ownership coverage in mind. These measures provide evidence about liquidity-related activity, but they do not directly measure market liquidity. The code and data used in the analysis are available at this https URL.