5 papers
Impact Is Not Invalidation: Ask About the Claim, Not the Diff
Atul Anand
Memory systems for coding agents must decide, when a repository changes, which of their stored claims have become false. Content anchoring invalidates a claim whenever the artifact…
Sequential knowledge editing breaks a model's ability to tell good evidence from bad, without costing it accuracy
Atul Anand
Knowledge editing is evaluated on whether the edited fact changed, whether paraphrases follow, and whether unrelated answers stayed put. A model can pass all three and still lose s…
Ask Which, Not How Good: Sizing Benchmarks Scored by an LLM
Atul Anand
Benchmarks scored by an LLM judge routinely adjudicate differences of a tenth of a point, but the resolution of those benchmarks has never been measured. Existing sample-complexity…
Diagnosing Tool-Selection Reasoning in LLM Agents with Canary Tools
Atul Anand, Sourav Chattaraj
Agent evaluations tell us that a model picked the wrong tool, but rarely why. We introduce canary tools: diagnostic probe tools planted in an agent's Model Context Protocol (MCP) t…
Instruction Stacking Collapse: A Benchmark and the Capability-Dependent Value of Prompt Compilation
Atul Anand, Sourav Chattaraj
Production prompts rarely carry a single instruction. One system message may require valid JSON, a word limit, three citations, and a fixed tone at the same time. We study how inst…