Croissant, a JSON‑LD based descriptor for machine‑learning datasets, now includes a policy profile that specifies how data‑use conditions are evaluated. The profile defines five operators with explicit decision procedures, enabling a gate to determine admissible operations from the descriptor alone and record what was checked. Experiments on two corpora and a real nf‑core pipeline show that the profile’s decisions match native descriptors, add minimal overhead, and cover all operators, refusal classes, and conformance clauses.
arXiv:2608. 09028v1 Announce Type: new Abstract: Institutional policies stay in natural language while the systems that check compliance demand machine-readable constraints.
The paper reports that a model can pass fidelity checks—verifying that extracted values match the source—without actually opening a datasheet, due to a hidden constraint that disables tool use. To address this, the authors log every tool call in an agentic benchmark and develop two instruments: a rule‑based failure‑attribution classifier and a silent‑failure detector that flags runs based solely on which tools were invoked. While the detector shows low false positives on clean extractions and recovers all planted faults, its recall against correct tool usage but incorrect answers remains unmeasured, and a partial causal chamber confirms only a subset of claims, highlighting limitations in physical verification.
arXiv:2609.06036v1 Announce Type: new Abstract: Proposal-based controllers---learned policies, language-model planners, and other black-box \emph{generators}---are increasingly deployed behind runtim...
arXiv:2608. 15810v1 Announce Type: new Abstract: Runtime compression of serving state trades quality for capacity with no priced guarantee: systems adapt precision on load signals with no soundness statement, and certified approaches budget request-level risk by a union bound over a pre-declared event count.
arXiv:2609.22259v1 Announce Type: new Abstract: Context layers, curated documentation that an analytics agent fetches at query time, produce large accuracy gains on text-to-SQL benchmarks. A with/wit...
arXiv:2608. 16852v1 Announce Type: new Abstract: Regulatory compliance monitoring in deployed language models is increasingly implemented as a legal and audit control, checking model outputs against written rules spanning data protection, healthcare, financial regulation, and platform policy.
arXiv:2606. 07316v2 Announce Type: replace-cross Abstract: Can a committee of LLM agents reach agreement that is certifiable at the level of meaning, not only at the level of a label?
arXiv:2608. 07946v1 Announce Type: cross Abstract: Text-to-SQL benchmarks ship schemas whose column names already say what the columns mean.
arXiv:2606. 29091v1 Announce Type: cross Abstract: Tabular foundation models cannot reason about data produced by running systems without access to the rules that govern them.
arXiv:2607. 23386v1 Announce Type: new Abstract: We document a failure class in frontier large language models -- exception chain collapse -- observed in eligibility evaluation under nested conditional rules of the form "A is required UNLESS B applies, UNLESS C overrides B".
arXiv:2607. 25364v1 Announce Type: new Abstract: Tool-using agents expose structured calls but commonly attach free-form rationales.
arXiv:2607. 10069v1 Announce Type: new Abstract: Semantic caching defines answer reuse on embedding similarity: two utterances share a stored answer when a similarity score clears a threshold, with no notion of authorization, versioning, or of what makes two demands the same.