5 papers
The Refusal Residue: When Probes Catch Alignment Faking and When They Don't
Aman Mehta
Alignment faking is dangerous because a model can appear compliant under monitoring while preserving behavior it would reveal when unmonitored. When no scratchpad is visible, behav…
Finding the Right Tables and Columns: A Benchmark and Corpus-Adaptive Embeddings for SQL Schema Retrieval
Qingcheng Zeng, Puxuan Yu, Aman Mehta +2
Retrieval in the SQL setting has largely been studied as the task of finding, within a large collection of SQL statements, the statement that answers a natural-language question. A…
Plans Don't Persist: Why Context Management Is Load Bearing for LLM Agents
Aman Mehta, Anupam Datta
Long-horizon agents depend on context management: systems compress, summarize, and evict old tokens so tasks can continue beyond finite windows. That is safe only when dropped info…
When Agents Commit Too Soon: Diagnosing Premature Commitment in LLM Agents
Aman Mehta
Long-horizon LLM agents can fail quietly: they settle on one reading of the evidence early, then spend the rest of the run defending it. We call this premature commitment. Final-an…
Dual-View Training for Instruction-Following Information Retrieval
Qingcheng Zeng, Puxuan Yu, Aman Mehta +2
Instruction-following information retrieval (IF-IR) studies retrieval systems that must not only find documents relevant to a query, but also obey explicit user constraints such as…