Asia
Aligning Gradient and Hessian for Neural Signed Distance Function
Our motivation is grounded in a fundamental observation: aligning the gradient and the Hessian of the SDF provides a more efficient mechanism to govern gradient directions.
Mixture of Demonstrations for In-Context Learning Song Wang
We further design an expert-wise training strategy to alleviate the impact of unhelpful demonstrations when optimizing the retriever model.
DeformableTST: Transformer for Time Series Forecasting without Over-reliance on Patching
In this paper, we intent to handle this emerging issue.
Understanding Transformer Predictions Through Memory Efficient Attention Manipulation
Most crucially, they require prohibitively large amounts of additional memory since they rely on backpropagation which allocates almost twice as much GPU memory as the forward pass. This renders it difficult, if not impossible, to use explanations in production.
Implicit variance regularization in non-contrastive SSL
In this work, we provide a comparative analysis of the learning dynamics for the Euclidean and cosine-based asymmetric losses in the eigenspace of the closed-form predictor DirectPred.
