10 papers
Visual Abstention in Unified Multimodal Models
Chufan Shi, Cheng Yang, Tiannuo Yang +4
Unified multimodal models (UMMs) integrate understanding and generation, yet their generative behavior is rarely governed by what they understand about the task. We formalize visua…
Towards Looped Models Done Right, Part II: Rethinking at Fixed Points
Benhao Huang, Chufan Shi, Junlin Chen +4
Every recurrence of a looped language model adds cost in training, decoding, prefill, and reinforcement learning (RL). The closer recurrent states get to fixed points, the less the…
Recursive Self-Improvement in Unified Multimodal Models
Huijuan Wang, Chufan Shi, Cheng Yang +3
Unified multimodal models (UMMs) understand and generate both text and images, which lets a model produce its own training data. Existing self-improvement in UMMs keeps supervision…
InfoEdit: Probing Global Layout Reasoning in Infographic Editing
Cheng Yang, Chufan Shi, Huijuan Wang +6
Multimodal foundation models edit natural photographs at production quality, yet the same models struggle with structured visual content such as infographics. Unlike photographs, i…
Do Text Edits Generalize to Visual Generation? Benchmarking Cross-Modal Knowledge Editing in UMMs
Xin Gao, Cheng Yang, Chufan Shi +1
Unified multimodal models (UMMs) have emerged as a promising paradigm for general-purpose multimodal intelligence. As they are deployed in real-world applications, effectively upda…
Are VLMs Seeing or Just Saying? Uncovering the Illusion of Visual Re-examination
Chufan Shi, Cheng Yang, Yaokang Wu +4
Vision-Language Models (VLMs) often produce self-reflective statements like "let me check the figure again" during reasoning. Do such statements trigger genuine visual re-examinati…
EMO: Frustratingly Easy Progressive Training of Extendable MoE
Linghao Jin, Chufan Shi, Huijuan Wang +4
Sparse Mixture-of-Experts (MoE) models offer a powerful way to scale model size without increasing compute, as per-token FLOPs depend only on k active experts rather than the total…
Asymmetric Idiosyncrasies in Multimodal Models
Muzi Tao, Chufan Shi, Huijuan Wang +2
In this work, we study idiosyncrasies in the caption models and their downstream impact on text-to-image models. We design a systematic analysis: given either a generated caption o…
UReason: Benchmarking Reasoning-to-Generation Alignment in Unified Multimodal Models
Cheng Yang, Chufan Shi, Bo Shui +7
Unified multimodal models (UMMs) aim to integrate multimodal understanding and generation within a unified architecture, yet it remains unclear to what extent textual and visual mo…
From Abstract to Contextual: What LLMs Still Cannot Do in Mathematics
Bowen Cao, Dongdong Zhang, Yixia Li +8
Large language models now solve many benchmark math problems at near-expert levels, yet this progress has not fully translated into reliable performance in real-world applications.…