Ravid Shwartz-Ziv

From Tokens to Thoughts: How LLMs and Humans Trade Compression for Meaning

Attention Sinks and Compression Valleys in LLMs are Two Sides of the Same Coin

Antislop: A Comprehensive Framework for Identifying and Eliminating Repetitive Patterns in Language Models

NdLinear: Preserving Multi-Dimensional Structure for Parameter-Efficient Neural Networks

A superpersuasive autonomous policy debating system

Turning Up the Heat: Min-p Sampling for Creative and Coherent LLM Outputs

LiveBench: A Challenging, Contamination-Limited LLM Benchmark

Layer by Layer: Uncovering Hidden Representations in Language Models

Seq-VCR: Preventing Collapse in Intermediate Transformer Representations for Enhanced Reasoning

Inheritune: Training Smaller Yet More Attentive Language Models

Just How Flexible are Neural Networks in Practice?

Variance-Covariance Regularization Improves Representation Learning