Information Theory

Model Compression and Efficient AI

Ravid Shwartz Ziv's research on model compression, task-aware quantization, efficient representations, and reducing AI memory and computation.

Layer by Layer: Uncovering Hidden Representations in Language Models

Finding useful embeddings inside language models. We study why intermediate layers can outperform the final layer and how to identify them using information, geometry, and downstream evaluation.

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

Comparing how language models and humans balance compact representations with semantic detail.

Rate-In: Information-Driven Adaptive Dropout Rates for Improved Inference-Time Uncertainty Estimation

Adjusting dropout during inference to estimate uncertainty without additional labels or training.

An Information Theory Perspective on Variance-Invariance-Covariance Regularization

We provide an information-theoretic analysis of VICReg, deriving theoretical foundations for deterministic networks and introducing new SSL methods based on these insights.

To Compress or Not to Compress--Self-Supervised Learning and Information Theory: A Review

We present a comprehensive review of self-supervised learning through the lens of information theory, introducing a unified framework that encompasses existing approaches and highlighting the interplay between compression and information preservation in deep neural networks.