2 papers
cs.LG2025
Data-Free Pruning of Self-Attention Layers in LLMs
Dhananjay Saikumar, Blesson Varghese
Many self-attention sublayers in large language models (LLMs) can be removed with little to no loss. We attribute this to the Attention Suppression Hypothesis: during pre-training,…
cs.LG2025
Signal Collapse in One-Shot Pruning: When Sparse Models Fail to Distinguish Neural Representations
Dhananjay Saikumar, Blesson Varghese
Neural network pruning is essential for reducing model complexity to enable deployment on resource constrained hardware. While performance loss of pruned networks is often attribut…