9 papers
Few-Step Diffusion Language Models via Trajectory Self-Distillation
Tunyu Zhang, Xinxi Zhang, Ligong Han +9
Diffusion large language models (DLLMs) have emerged as powerful generative models with the promise of fast text generation through parallel decoding. However, realizing this poten…
Beyond Interpretability: When, Why, and How Sparse Autoencoders Enable Label-Free Visual Steering
Gerasimos Chatzoudis, Zhuowei Li, Gemma E. Moran +2
Sparse Autoencoders (SAEs) are increasingly used to interpret foundation models, but their role as an actionable intervention space remains less understood, especially in vision. W…
Can Cross-Layer Transcoders Replace Vision Transformer Activations? An Interpretable Perspective on Vision
Gerasimos Chatzoudis, Konstantinos D. Polyzos, Zhuowei Li +4
Understanding the internal activations of Vision Transformers (ViTs) is critical for building interpretable and trustworthy models. While Sparse Autoencoders (SAEs) have been used…
TokUR: Token-Level Uncertainty Estimation for Large Language Model Reasoning
Tunyu Zhang, Haizhou Shi, Yibin Wang +9
While Large Language Models (LLMs) have demonstrated impressive capabilities, their output quality remains inconsistent across various application scenarios, making it difficult to…
Decoupling Vision and Language: Codebook Anchored Visual Adaptation
Jason Wu, Tianchen Zhao, Chang Liu +7
Large Vision-Language Models (LVLMs) use their vision encoders to translate images into representations for downstream reasoning, but the encoders often underperform in domain-spec…
The Hidden Life of Tokens: Reducing Hallucination of Large Vision-Language Models via Visual Information Steering
Zhuowei Li, Haizhou Shi, Yunhe Gao +7
Large Vision-Language Models (LVLMs) can reason effectively over both textual and visual inputs, but they tend to hallucinate syntactically coherent yet visually ungrounded content…