AI Paper Atlas

Research map / Large language models / Reasoning and chain of thought

DDCoT: Duty-Distinct Chain-of-Thought Prompting for Multimodal Reasoning in Language Models

Ge Zheng, Bin Yang, Jiajin Tang, Hong-Yu Zhou, Sibei Yang

NeurIPS 2023 · Large language models › Reasoning and chain of thought · approach: Chain-of-thought training and analysis · 21 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
Use DDCoT instead of GPT-3 (CoT)GPT-3 on ScienceQA · ScienceQA test, IMGaccuracy: 67.43 -> 69.96Improved
Use DDCoT instead of ChameleonChatGPT on ScienceQA · ScienceQA test, IMGaccuracy: 70.80 -> 72.53Improved
Use DDCoT instead of MM-CoTUnifiedQA on ScienceQA · ScienceQA test, Avgaccuracy: +2.43%Improved
Add DDCoT rationalesFlamingo on ScienceQA · ScienceQA test, Avgaccuracy: 27.87 -> 39.01Improved
Add DDCoT rationalesMiniGPT-4 on ScienceQA · ScienceQA test, Avgaccuracy: 44.71 -> 55.67Improved
Add captions to the language-only baseline (ablation)T5 baseline on ScienceQA · ScienceQA test, IMGaccuracy: 72.93 -> 75.26Improved
Use original image input without additional components (ablation)T5 baseline on ScienceQA · ScienceQA test, IMGaccuracy: 72.93 -> 62.82Hurt
Add DDCoT rationales instead of using no rationale (ablation)T5 baseline on ScienceQA · ScienceQA test, IMGaccuracy: 75.16 -> 83.34Improved
Replace naive rationale with decomposed-question rationale (ablation)DDCoT rationale generation · ScienceQA test, IMGaccuracy: 75.06 -> 75.61Improved
Add duty-distinct prompting with negative-space uncertainty (ablation)DDCoT rationale generation · ScienceQA test, IMGaccuracy: 75.61 -> 83.34Improved
Add explicit uncertainty prompting (ablation)Duty-distinct rationale generation · ScienceQA test, IMGaccuracy: 78.19 -> 83.34Improved
Add joint reasoning integration (ablation)DDCoT rationale generation · ScienceQA test, IMGaccuracy: 77.49 -> 83.34Improved

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