AI Paper Atlas

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

ReasonEdit: Editing Vision--Language Models using Human Reasoning

Jiaxing Qiu, Kaihua Hou, Roxana Daneshjou, Ahmed Alaa, Thomas Hartvigsen

ICML 2026 · Vision-language and multimodal › Visual question answering · approach: Vision-language model VQA evaluation · 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 reasoning entries to the codebook instead of answer entries only (ablation)ReasonEdit · Entry-type ablation, LLaVA-1.5-7BR-Gen: 0.65 -> 0.87Improved
add reasoning entries to the codebook instead of answer entries only (ablation)ReasonEdit · Entry-type ablation, LLaVA-1.5-7BCoE-Gen: 0.72 -> 0.82Improved
replace ReasonEdit by best prior editor with reasoning (IKE-COT) on R-GenReasonEdit · FVQA, LLaVA-1.5-7BR-Gen: 0.81 -> 0.87Improved
use ReasonEdit instead of IKE-COTReasonEdit · FVQA, Qwen3-VL-8BCoE-Gen: 0.77 -> 0.95Improved
add human reasoning supervision (COT) to FTFT · FVQA, LLaVA-1.5-7BR-Gen: 0.39 -> 0.61Improved
add human reasoning (COT) to GRACEGRACE · FVQA, LLaVA-1.5-7BCoE-Gen: 0.54 -> 0.54No clear effect
add human reasoning (COT) to MENDMEND · FVQA, LLaVA-1.5-7BAcc: 0.52 -> 0.45Hurt
inject noisy statements (50%) into reasoning (ablation)ReasonEdit · Noisy reasoning, LLaVA-1.5-7BR-Gen: 0.86 -> 0.82No clear effect
replace 50% of reasoning facts with noise (ablation)ReasonEdit · Noisy reasoning, LLaVA-1.5-7BR-Gen: 0.86 -> 0.77Hurt
replace 50% of reasoning facts with noise (ablation)ReasonEdit · Noisy reasoning, LLaVA-1.5-7BCoE-Gen: 0.98 -> 0.93Hurt
add dual vision-language embedding instead of a single vision layer key (ablation)ReasonEdit · codebook keysimage generalityImproved
add dual embedding instead of single language layer key (ablation)ReasonEdit · codebook keystext generalityImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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