arXiv:2608. 14992v1 Announce Type: new Abstract: Language-model systems increasingly read from stores they also write to, so a claim that was merely written earlier can return looking retrieved.
arXiv:2604. 16706v2 Announce Type: replace Abstract: Automated evaluation of tool-using large language model (LLM) agents is widely assumed to be reliable, yet this assumption is rarely validated against human annotation.
arXiv:2608. 07946v1 Announce Type: cross Abstract: Text-to-SQL benchmarks ship schemas whose column names already say what the columns mean.
arXiv:2606. 27383v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used as research assistants, yet it remains unclear whether they can calibrate research takeaways to the strength and scope of the supporting evidence.
arXiv:2606. 30689v1 Announce Type: cross Abstract: Spec-Driven Development (SDD) frameworks guide Large Language Model (LLM)-powered code generation through formal specifications, yet they differ fundamentally in how they enforce traceability between requirements and generated code.
arXiv:2608. 04278v1 Announce Type: cross Abstract: Coding agents increasingly work across sessions, but prose notes can preserve a conclusion without the program state that supported it.
arXiv:2607. 19257v1 Announce Type: cross Abstract: Practitioners make three prompt-design decisions with almost no controlled evidence behind them: how to format instructions and context (markdown, plain text, prose, or tabular), how many simultaneous instructions a system prompt can carry before compliance degrades, and how much context a model can hold before recall and honesty degrade.
arXiv:2605. 28734v2 Announce Type: replace-cross Abstract: A general-purpose language model that answers a harmful question returns text; a coding model that complies with a malicious request can return a working weapon: a keylogger, ransomware, an exploit that runs as written.
arXiv:2607. 25364v1 Announce Type: new Abstract: Tool-using agents expose structured calls but commonly attach free-form rationales.
arXiv:2607. 09691v1 Announce Type: cross Abstract: A modern coding agent can hold an entire repository in its context window.
arXiv:2607. 14399v1 Announce Type: new Abstract: Evaluations of language-model honesty read the model's verdicts as evidence about the model.
Practitioners make three prompt-design decisions with almost no controlled evidence behind them: how to format instructions and context (markdown, plain text, prose, or tabular), how many simultaneous instructions a system prompt can carry before compliance degrades, and how much context a model can hold before recall and honesty degrade. We report two controlled experiments crossing all three factors on one held, contamination-free synthetic corpus (the "Book of Veyra," 8,780 uniquely-named entities, deterministically regenerable from a fixed seed), evaluated across five models.
arXiv:2607. 20827v1 Announce Type: new Abstract: LLM agents choose tools and arguments from context that mixes user requests, tool outputs, retrieved records, memory, and untrusted text.