PSEBench - Missing Slot Identification: leaderboard
A PSEBench patient safety triage dimension: the F1 of the withheld fields that the model questions target, on missing-information cases where it asked, in percent; higher is better. PSEBench (2026) builds 5,074 verifiable cases from clause cards of the Minnesota 29 reportable adverse health events, and models may query a simulated information provider before answering.
Metric: Missing slot identification F1 (%, M6): micro F1 of the withheld fields the ASK queries target, on missing cases where the model asked, on the 5,074-case MN29 benchmark (3,455 complete, 1,362 missing-information and 257 uncertain cases built from clause cards of the Minnesota 29 reportable adverse health events), agentic environment where the model may ASK a GPT-5.2 information provider within 10 turns and then answers with a verdict, clause, legal evidence and rationale; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 77.6 | #79 |
| 2 | Gemini 3.1 Pro (Preview) | 77.4 | #94 |
| 3 | GPT-5.5 | 77.3 | #49 |
| 4 | Claude Sonnet 4.6 | 76.2 | #161 |
| 5 | GPT-5 | 75 | #167 |
| 6 | Gemini 2.5 Flash | 71.6 | #453 |
| 7 | GPT-OSS-120B | 71.6 | #700 |
| 8 | Qwen 3 235B A22B 2507 Instruct | 69.3 | #619 |
| 9 | DeepSeek R1 | 66.3 | #473 |
| 10 | GPT-5 Nano | 61 | #889 |
| 11 | Mistral Small 3.2 | 53.5 | #1591 |
| 12 | Llama 3.1 8B Instruct | 24.6 | #4246 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=psebench-missing-slot-identification · How It Works · Data refreshed daily, snapshot 2026-11-13.