AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Evaluating Open-QA Evaluation

Cunxiang Wang, Sirui Cheng, Qipeng Guo, Yuanhao Yue, Bowen Ding, Zhikun Xu, Yidong Wang, Xiangkun Hu, Zheng Zhang, Yue Zhang

NeurIPS 2023 · Large language models › Retrieval and knowledge · approach: Explanatory and multiple-choice QA benchmarks · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Ask the evaluator to give reasons for its judgments (ablation)GPT-3.5 evaluator · EVOUNA NQ-BingChataccuracy/F1: 69.5/77.2 -> 55.6/62.2Hurt
Ask the evaluator to give reasons for its judgments (ablation)GPT-3.5 evaluator · EVOUNA NQ-ChatGPT4accuracy/F1: 80.9/86.9 -> 64.3/71.2Hurt
Tell the evaluator to ignore background information (ablation)GPT-3.5 evaluator · EVOUNA NQ-BingChataccuracy/F1: 69.5/77.2 -> 65.7/73.4Hurt
Tell the evaluator to ignore background information (ablation)GPT-3.5 evaluator · EVOUNA NQ-FiDaccuracy/F1: 93.6/95.3 -> 93.9/95.5No clear effect
Add chain-of-thought prompting (ablation)GPT-3.5 evaluator · EVOUNA NQ-BingChataccuracy/F1: 69.5/77.2 -> 80.4/87.1Improved
Add chain-of-thought prompting (ablation)GPT-3.5 evaluator · EVOUNA NQ-ChatGPT4accuracy/F1: 80.9/86.9 -> 86.0/91.2Improved
Add chain-of-thought prompting (ablation)GPT-3.5 evaluator · EVOUNA NQ-FiDaccuracy/F1: 93.6/95.3 -> 90.2/92.9Hurt
Add in-context learning examples (ablation)GPT-3.5 evaluator · EVOUNA NQ-BingChataccuracy/F1: 69.5/77.2 -> 75.3/82.5Improved
Add in-context learning examples (ablation)GPT-3.5 evaluator · EVOUNA NQ-ChatGPT4accuracy/F1: 80.9/86.9 -> 83.2/88.6Improved
Add in-context learning examples (ablation)GPT-3.5 evaluator · EVOUNA NQ-FiDaccuracy/F1: 93.6/95.3 -> 93.1/94.9No clear effect
Exclude the NQ-BingChat subset from evaluation (ablation)GPT-3.5 evaluator · EVOUNA-NQfalse negative proportion: 14.3 -> 9.9Improved
Use lexical matching instead of human judgment to score QA modelsOpen-QA evaluation · NaturalQuestions ChatGPT-4accuracy: 78.8 -> 61.8Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy