DialToM - Prospective (Motivational Interviewing): leaderboard
Metric: Correctness rate (%) on four-option multiple-choice questions of the DialToM Prospective task: forecast the state-consistent next dialogue trajectory from an isolated mental-state profile with no dialogue context (the State-Driven Diagnostic Probe), motivational interviewing domain (naturalistic human-human dialogues from AnnoMI; GPT-4o-drafted options verified by professional annotators and Dawid-Skene aggregation), zero-shot; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 83.09 |
| 2 | Gemini 2.5 Pro | 38.24 |
| 3 | Gemini 2.5 Flash | 26.47 |
| 4 | Kimi K2 | 22.06 |
| 5 | Llama 3.3 70B | 18.42 |
| 6 | Qwen 3 235B A22B | 17.65 |
| 7 | GPT-5 | 14.71 |
| 8 | Llama 4 Maverick | 9.56 |
| 9 | GPT-OSS-120B | 7.35 |
| 10 | Llama 3.1 8B | 6.62 |
| 11 | Mistral Nemo | 6.62 |
| 12 | GPT-4.1 | 5.88 |
| 13 | Mistral Small 3.2 | 5.88 |
Interactive version: theaggregate.ai/benchmark?slug=dialtom-prospective-motivational-interviewing · How It Works · Data refreshed daily, snapshot 2026-10-07.