Study: Can LLMs recognize missing info and ask questions?
marinkazitnik · x · 2026-08-19
Marinka Zitnik shared new research investigating whether LLMs know when they have enough information to answer, and what to ask when they don't.
- Core Problem: Most agent research focuses on answering, but the reverse ability—recognizing missing info and asking—is critical for high-stakes fields like healthcare.
- Example: In pelvic pain diagnosis, the pain duration dictates the pathway; without asking, downstream symptoms are irrelevant.
- Outcome: A paper and project were released to improve agent accuracy and safety in human-in-the-loop workflows.
More from Research
- Classic Paper: The Surprising Creativity of Digital Evolution — Ghost_Pilot_MD · 2026-08-19
- HarmProfile Benchmark: Harmfulness and Diversity Rise with Model Capability — Zhouyuan Ma · 2026-08-19
- Why Removing the Vision Encoder Can Be Better — From an Infra Perspective — liuziwei7 · 2026-08-19
- Matmul Optimization Bottleneck: Data Movement, Not Multiplications — yaroslavvb · 2026-08-19
- Hot take: "Skill optimization" is just repackaged prompt optimization — IanArawjo · 2026-08-19
- NBER study: Pangram detector achieves near-zero error rates — soumitrashukla9 · 2026-08-19