8 papers
Human-Centric Image Captioning with Subject-Centered Spatial Understanding
Bozhou Li, Jiahang Zhang, Yue Ding +11
While multimodal large language models (MLLMs) achieve remarkable performance on generic image captioning, they frequently suffer from structural hallucinations in human-centric sc…
Flux-OPD: On-Policy Distillation with Evolving Contexts
Yuran Wang, Zekun Wang, Bohan Zeng +10
Large language model training in open-ended domains lacks verifiable rewards, making task preferences difficult to formalize as effective supervision. Contexts can convey such pref…
ProphetKV: User-Query-Driven Selective Recomputation for Efficient KV Cache Reuse in Retrieval-Augmented Generation
Shihao Wang, Jiahao Chen, Yanqi Pan +7
The prefill stage of long-context Retrieval-Augmented Generation (RAG) is severely bottlenecked by computational overhead. To mitigate this, recent methods assemble pre-calculated…
Semantic Routing: Exploring Multi-Layer LLM Feature Weighting for Diffusion Transformers
Bozhou Li, Yushuo Guan, Haolin Li +7
Recent DiT-based text-to-image models increasingly adopt LLMs as text encoders, yet text conditioning remains largely static and often utilizes only a single LLM layer, despite pro…
CoF-T2I: Video Models as Pure Visual Reasoners for Text-to-Image Generation
Chengzhuo Tong, Mingkun Chang, Shenglong Zhang +12
Recent video generation models have revealed the emergence of Chain-of-Frame (CoF) reasoning, enabling frame-by-frame visual inference. With this capability, video models have been…
The Unseen Bias: How Norm Discrepancy in Pre-Norm MLLMs Leads to Visual Information Loss
Bozhou Li, Xinda Xue, Sihan Yang +5
Multimodal Large Language Models (MLLMs), which couple pre-trained vision encoders and language models, have shown remarkable capabilities. However, their reliance on the ubiquitou…