AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

UniGame: Turning a Unified Multimodal Model Into Its Own Adversary

Zhaolong Su, Wang Lu, Hao Chen, Sharon Li, Jindong Wang

CVPR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities · 8 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 UniGame self-adversarial post-trainingJanus-Pro-7B · Consistency (UnifiedBench + WISE)consistency score: 63.66 -> 68.32Improved
use UniGame instead of plain SFTJanus-Pro-7B · Consistency (UnifiedBench + WISE)consistency score: 64.72 -> 68.32Improved
add UniGame self-adversarial post-trainingJanus-Pro-7B · Understanding overallaverage score: 71.4 -> 75.0Improved
use UniGame instead of plain SFTJanus-Pro-7B · Understanding overallaverage score: 71.9 -> 75.0Improved
add plain SFT post-trainingJanus-Pro-7B · Understanding overallaverage score: 71.4 -> 71.9No clear effect
add UniGame self-adversarial post-trainingJanus-Pro-7B · VQAv2 (test)accuracy: 78.2 -> 83.4Improved
add UniGame self-adversarial post-trainingJanus-Pro-7B · POPEaccuracy: 87.4 -> 89.6Improved
add UniGame self-adversarial post-trainingJanus-Pro-7B · GenEvaloverall score: 0.80 -> 0.82Improved
add plain SFT post-trainingJanus-Pro-7B · GenEvaloverall score: 0.80 -> 0.81No clear effect
UniGame (trained mainly on understanding) versus RecA with explicit generation post-trainingJanus-Pro-7B · GenEvaloverall score: 0.82 vs. 0.86Hurt
add UniGame self-adversarial post-trainingJanus-Pro-7B · NaturalBench (OOD robustness)group accuracy: +4.8%Improved
add UniGame self-adversarial post-trainingJanus-Pro-7B · AdVQA (adversarial robustness)accuracy: +6.2%Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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