paper-with-me

홈 › Papers

Bridging Information-Theoretic and Geometric Compression in Language Models

2023-10-20 · Emily Cheng, Corentin Kervadec, Marco Baroni

For a language model (LM) to faithfully model human language, it must compress vast, potentially infinite information into relatively few dimensions. We propose analyzing compression in (pre-trained) LMs from two points of view: geometric and information-theoretic. We demonstrate that the two views are highly correlated, such that the intrinsic geometric dimension of linguistic data predicts their coding length under the LM. We then show that, in turn, high compression of a linguistic dataset predicts rapid adaptation to that dataset, confirming that being able to compress linguistic information is an important part of successful LM performance. As a practical byproduct of our analysis, we evaluate a battery of intrinsic dimension estimators for the first time on linguistic data, showing that only some encapsulate the relationship between information-theoretic compression, geometric compression, and ease-of-adaptation.

📄 PDF Abstract BibTeX arXiv:2310.13620

Code (1)

chengemily1/id_bridging 공식 구현 pytorch

Tasks

Language ModelingLanguage Modelling

Similar Papers 제목 키워드 기반

Geometric and Information Compression of Representations in Deep Learning

2026-06-19 · Linara Adilova, Henning Petzka, Asja Fischer, Bernhard C. Geiger arxiv

Deep neural networks transform input data into latent representations that support a wide range of downstream tasks. These representations can be characterized along information-theoretic and geometric dimensions, but th…

Large Language Models for Lossless Image Compression: Next-Pixel Prediction in Language Space is All You Need

2024-11-19 · Kecheng Chen, Pingping Zhang, Hui Liu, Jie Liu 외

We have recently witnessed that ``Intelligence" and `` Compression" are the two sides of the same coin, where the language large model (LLM) with unprecedented intelligence is a general-purpose lossless compressor for va…

AllAttributeImage Compression

CLaRa: Bridging Retrieval and Generation with Continuous Latent Reasoning

2025-11-24 · Jie He, Richard He Bai, Sinead Williamson, Jeff Z. Pan 외 arxiv

Retrieval-augmented generation (RAG) enhances large language models (LLMs) with external knowledge but still suffers from long contexts and disjoint retrieval-generation optimization. In this work, we propose CLaRa (Cont…

Question Answering

On Information Plane Analyses of Neural Network Classifiers -- A Review

2020-03-21 · Bernhard C. Geiger

We review the current literature concerned with information plane analyses of neural network classifiers. While the underlying information bottleneck theory and the claim that information-theoretic compression is causall…

Information PlaneMutual Information Estimation

CoreMem: Riemannian Retrieval and Fisher-Guided Distillation for Long-Term Memory in Dialogue Agents

2026-06-16 · Jiaqi Chen, Yongqin Zeng, Shaoshen Chen, Yijian Zhang 외 arxiv

Personalized dialogue agents require continuous long-term memory to maintain coherent interactions across multiple sessions. However, deploying these capabilities on consumer-grade hardware (e.g., 8 GB VRAM edge devices)…