Deep Learning
Flow Priors for Linear Inverse Problems via Iterative Corrupted Trajectory Matching
Tweedie's formula, we show that we can perform gradient steps to sequentially optimize these objectives.
Spuriosity Rankings: Sorting Data to Measure and Mitigate Biases Mazda Moayeri 1 Wenxiao Wang
Work carried out while at the University of Maryland.
Vript: A Video Is Worth Thousands of Words Dongjie Y ang
As a video has an additional temporal dimension, it contains more information than a single image.
A Unified Pixel-level Vision LLM for Understanding, Generating, Segmenting, Editing Hao Fei 1,2 Shengqiong Wu
Shuicheng Y an is the corresponding author.
Inference-Time Intervention: Eliciting Truthful Answers from a Language Model Kenneth Li
This intervention significantly improves the performance of LLaMA models on the TruthfulQA benchmark.
Enhancing Graph Transformers with Hierarchical Distance Structural Encoding Y uankai Luo 1,3 Hongkang Li2 Lei Shi 1 Xiao-Ming Wu3 1
MPNNs, graph transformers do not rely on fixed graph structure information.