CLAMBER - Zero-shot CoT - Clarification Helpfulness: leaderboard
CLAMBER (ACL 2024) evaluates detecting ambiguous information needs and generating clarifying questions. Zero-shot CoT - Clarification Helpfulness under the named prompting protocol. Metric: Helpfulness (%). Helpfulness is human-rated; BERTScore measures similarity to reference questions and is not task success. Source: arXiv 2405.12063v2, tables 3 and 7. Chat and instruct models remain distinct; ChatGPT here is GPT-3.5-Turbo-16k.
Metric: Helpfulness (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | vicuna-13B | 40.97 | #1450 |
| 2 | Llama 2 13B Chat | 40.4 | #1414 |
| 3 | Llama 2 70B Chat | 32.72 | #1275 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=clamber-zero-shot-cot-clarification-helpfulness · How It Works · Data refreshed daily, snapshot 2026-11-05.