CVPR Findings2026
AEGIS: Exploring the Limit of World Knowledge Capabilities for Unified Multimodal Models
Jintao Lin, Bowen Dong, Weikang Shi, Chenyang Lei, Suiyun Zhang, Rui Liu, Xihui Liu
Abstract
[Image] Is the square g1, which was originally occupied by a white knight, now empty in the edited image? … [Image] Does the edited image correctly show a black pawn on the f6 square, consistent with the original board state?" … Question <image1> is the classic Italian Game opening in chess. Please illustrate the position after one half-move.
Gemini
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- Show-o2: Improved Native Unified Multimodal Models cites MMaDA: Multimodal Large Diffusion Language Models
- MMaDA: Multimodal Large Diffusion Language Models cites SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis
Builds on21
- SDXL: Improving Latent Diffusion Models for High-Resolution Image SynthesisDustin Podell, Zion English, Kyle Lacey, Andreas Blattmann et al.ICLR 2024 · 4,569 citations
- GAIA: a benchmark for General AI AssistantsGrégoire Mialon, Clémentine Fourrier, Thomas Wolf, Yann LeCun et al.ICLR 2024 · 716 citations
- Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion ModelsShihao Zhao, Dongdong Chen, Yen-Chun Chen, Jianmin Bao et al.NeurIPS 2023 · 505 citations
- Show-o2: Improved Native Unified Multimodal ModelsJinheng Xie, Zhenheng Yang, Mike Zheng ShouNeurIPS 2025 · 261 citations
- MMaDA: Multimodal Large Diffusion Language ModelsLing Yang, Ye Tian, Bowen Li, Xinchen Zhang et al.NeurIPS 2025 · 255 citations
Related papers
- UniGen-1.5: Enhancing Image Generation and Editing through Reward Unification in RLRui Tian, Mingfei Gao, Haiming Gang, Jiasen Lu et al.CVPR 2026
- Chessformer: A Unified Architecture for Chess ModelingDaniel Monroe, George Eilender, Philip Chalmers, Zhenwei Tang et al.ICLR 2026 · 8 citations
- Verification of the Implicit World Model in a Generative Model via Adversarial SequencesAndrás Balogh, Márk JelasityICLR 2026 · 1 citation
- General search techniques without common knowledge for imperfect-information games, and application to superhuman Fog of War chessBrian Zhang, Tuomas SandholmICLR 2026 · 13 citations
- Enhanced Visual Instruction Tuning with Synthesized Image-Dialogue DataYanda Li, Chi Zhang, Gang Yu, Wanqi Yang et al.ACL Findings 2024 · 8 citations
