5 papers
RobuQ: Pushing DiTs to W1.58A2 via Robust Activation Quantization
Kaicheng Yang, Xun Zhang, Haotong Qin +4
Diffusion Transformers (DiTs) have recently emerged as a powerful backbone for image generation, demonstrating superior scalability and performance over U-Net architectures. Howeve…
SRA 2: Variational Autoencoder Self-Representation Alignment for Efficient Diffusion Training
Mengmeng Wang, Dengyang Jiang, Liuzhuozheng Li +6
Denoising-based diffusion transformers, despite their strong generation performance, suffer from inefficient training convergence. Existing methods addressing this issue, such as R…
Attention Debiasing for Token Pruning in Vision Language Models
Kai Zhao, Wubang Yuan, Yuchen Lin +5
Vision-language models (VLMs) typically encode substantially more visual tokens than text tokens, resulting in significant token redundancy. Pruning uninformative visual tokens is…
A Good Plan is Hard to Find: Aligning Models with Preferences is Misaligned with What Helps Users
Nishant Balepur, Matthew Shu, Yoo Yeon Sung +5
To assist users in complex tasks, LLMs generate plans: step-by-step instructions towards a goal. While alignment methods aim to ensure LLM plans are helpful, they train (RLHF) or e…
Information-Guided Identification of Training Data Imprint in (Proprietary) Large Language Models
Abhilasha Ravichander, Jillian Fisher, Taylor Sorensen +6
High-quality training data has proven crucial for developing performant large language models (LLMs). However, commercial LLM providers disclose few, if any, details about the data…