AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

DIVA: Harnessing the Representation Divergence in Unified Multimodal Models for Mutual Reinforcement

Renjie Lu, Xulong Zhang, Xiaoyang Qu, Jianzong Wang, Shangfei Wang

ICML 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Vision-language pretraining · 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 DIVA post-trainingNexus-Gen · MMMUaccuracy: 43.5 -> 49.4Improved
add DIVA post-trainingShow-o · DPG-Benchscore: 69.81 -> 76.03Improved
add DIVA post-trainingLiquid · GenEvalscore: 0.70 -> 0.81Improved
plain SFT on the same post-training data instead of DIVA (ablation)Show-o · MMMUaccuracy: 26.3 -> 26.8No clear effect
plain SFT on same data (ablation)Show-o · GenEvalscore: 0.69 -> 0.67Hurt
add the unique-information regularizer Iuni (ablation)DIVA · Show-o MMMUaccuracy: 28.3 -> 32.4Improved
add stop-gradient in shared MI alignment (ablation)DIVA · Show-o DPG-Benchscore: 74.92 -> 76.03Improved
use gated MLP encoders instead of Linear+LN (ablation)DIVA · Show-o MMMUaccuracy: 29.4 -> 32.4Improved
use Transformer encoders instead of gated MLP (ablation)DIVA · Show-oMMMU/POPE/DPG-Bench: 32.4 -> 32.1No clear effect
use random masks instead of contiguous block mask (ablation)DIVA · Show-o MMMUaccuracy: 24.7 -> 32.4Improved
vary middle-layer range (9-17, 7-17, 7-19 vs 8-18) (ablation)DIVA · Show-oMMMU/GenEval/DPG-Bench: GenEval 0.72-0.75No clear effect
increase post-training data from 70K to 200K (ablation)DIVA · Show-o DPG-Benchscore: 75.83 -> 76.03Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 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