AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

Collaborative Enhancement of Large and Small Models for Question Answering via Dual Knowledge Transfer

Shaofei Wang, Yunan Liu, Xiaolan Tang, Wenlong Chen

AAAI 2026 · Vision-language and multimodal › Visual question answering · approach: Attention and prior-robust VQA · 0 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
add confidence-aware prompting (ablation)DKT large model (DeepSeek r1-14B) · ReCloraccuracy: 71.8 -> 73.6Improved
add distillation substep (ablation)DKT small model · ReCloraccuracy: 57.6 -> 59.6Improved
add judgment and reflection substeps (ablation)DKT small model · ReCloraccuracy: 57.8 -> 59.6Improved
add iteration between small4large and large4small (ablation)DKT small model · ReCloraccuracy: 59.0 -> 59.6Improved
add iteration between small4large and large4small (ablation)DKT large model · ReCloraccuracy: 72.8 -> 73.6Improved
add distillation substep (ablation)DKT small model · LogiQAaccuracy: 34.25 -> 37.33Improved
add confidence-aware prompting (ablation)DKT large model (DeepSeek r1-14B) · LogiQAaccuracy: 62.98 -> 64.67Improved
apply DKT to the small model with LLaMA 3-70B as large modelRoBERTa-based small model · CSQAaccuracy: 60.69 -> 63.96Improved
apply DKT to the small model with DeepSeek r1-14B as large model, even when large model is worsesmall model · ProtoQAMax Answers@10: 49.1 -> 51.7Improved
apply DKT large4small and small4large to LLaMA 3-70BLLaMA 3-70B · ReCloraccuracy: 62.6 -> 66.0Improved
apply DKT to LLaMA 3-70BLLaMA 3-70B · ProtoQAMax Answers@10: 38.2 -> 40.2No clear effect
run more than two iterations of transfer (ablation)DKT · LogiQA and ReCloraccuracyNo clear effect

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