# Semantic Compression With Large Language Models arXiv: https://arxiv.org/abs/2304.12512 · HTML: https://mlsanity.com/paper/2304.12512/ ## Key facts - Cited by 17 papers in the map (data through 14 Oct 2026). - Builds on 17 of its 40 references that are in the map. - First posted on arXiv 25 Apr 2023. ## Builds on (most cited first, 17 of 17) - [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021) - [Efficient Estimation of Word Representations in Vector Space](https://mlsanity.com/paper/1301.3781/) (2013) - [On the Opportunities and Risks of Foundation Models](https://mlsanity.com/paper/2108.07258/) (2021) - [Sparks of Artificial General Intelligence: Early experiments with GPT-4](https://mlsanity.com/paper/2303.12712/) (2023) - [Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing](https://mlsanity.com/paper/2107.13586/) (2021) - [A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity](https://mlsanity.com/paper/2302.04023/) (2023) - [A Prompt Pattern Catalog to Enhance Prompt Engineering with ChatGPT](https://mlsanity.com/paper/2302.11382/) (2023) - [Asleep at the Keyboard? Assessing the Security of GitHub Copilot's Code Contributions](https://mlsanity.com/paper/2108.09293/) (2021) - [code2vec: Learning Distributed Representations of Code](https://mlsanity.com/paper/1803.09473/) (2018) - [GitHub Copilot AI pair programmer: Asset or Liability?](https://mlsanity.com/paper/2206.15331/) (2022) - [Lost at C: A User Study on the Security Implications of Large Language Model Code Assistants](https://mlsanity.com/paper/2208.09727/) (2022) - [An Introduction to Neural Data Compression](https://mlsanity.com/paper/2202.06533/) (2022) - [Language Model Evaluation Beyond Perplexity](https://mlsanity.com/paper/2106.00085/) (2021) - [Is GitHub's Copilot as Bad as Humans at Introducing Vulnerabilities in Code?](https://mlsanity.com/paper/2204.04741/) (2022) - [Conversational Automated Program Repair](https://mlsanity.com/paper/2301.13246/) (2023) - [DECORE: Deep Compression with Reinforcement Learning](https://mlsanity.com/paper/2106.06091/) (2021) - [Evaluating Computational Language Models with Scaling Properties of Natural Language](https://mlsanity.com/paper/1906.09379/) (2019) ## Built on by (most cited first, 17 of 17) - [Large Language Models for Software Engineering: A Systematic Literature Review](https://mlsanity.com/paper/2308.10620/) (2023) - [LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models](https://mlsanity.com/paper/2310.05736/) (2023) - [TrustGPT: A Benchmark for Trustworthy and Responsible Large Language Models](https://mlsanity.com/paper/2306.11507/) (2023) - [Large Generative AI Models for Telecom: The Next Big Thing?](https://mlsanity.com/paper/2306.10249/) (2023) - [LeanContext: Cost-Efficient Domain-Specific Question Answering Using LLMs](https://mlsanity.com/paper/2309.00841/) (2023) - [At the Dawn of Generative AI Era: A Tutorial-cum-Survey on New Frontiers in 6G Wireless Intelligence](https://mlsanity.com/paper/2402.18587/) (2024) - [Automatically Generating UI Code from Screenshot: A Divide-and-Conquer-Based Approach](https://mlsanity.com/paper/2406.16386/) (2024) - [GenAINet: Enabling Wireless Collective Intelligence via Knowledge Transfer and Reasoning](https://mlsanity.com/paper/2402.16631/) (2024) - [Divide, Reweight, and Conquer: A Logit Arithmetic Approach for In-Context Learning](https://mlsanity.com/paper/2410.10074/) (2024) - [Parse Trees Guided LLM Prompt Compression](https://mlsanity.com/paper/2409.15395/) (2024) - [Steering Language Generation: Harnessing Contrastive Expert Guidance and Negative Prompting for Coherent and Diverse Synthetic Data Generation](https://mlsanity.com/paper/2308.07645/) (2023) - [Exploring the Limits of Semantic Image Compression at Micro-bits per Pixel](https://mlsanity.com/paper/2402.13536/) (2024) - [From Generalist to Specialist: Improving Large Language Models for Medical Physics Using ARCoT](https://mlsanity.com/paper/2405.11040/) (2024) - [SMIC: Semantic Multi-Item Compression based on CLIP dictionary](https://mlsanity.com/paper/2412.05035/) (2024) - [SemanticZip: A Pilot Framework for Lossy Text Compression with LLMs as Semantic Decompressors](https://mlsanity.com/paper/2605.24541/) (2026) - [An Exploratory Evaluation of LLM-Assisted Rewriting of Moderate-Complexity Financial Sentences for DisCoCat-Based Sentiment Analysis](https://mlsanity.com/paper/2608.07439/) (2026) - [Investigating the Influence of Prompt and Response Languages on LLM Content Generation](https://mlsanity.com/paper/2608.26186/) (2026) ## Abstract (arXiv) The rise of large language models (LLMs) is revolutionizing information retrieval, question answering, summarization, and code generation tasks. However, in addition to confidently presenting factually inaccurate information at times (known as "hallucinations"), LLMs are also inherently limited by the number of input and output tokens that can be processed at once, making them potentially less effective on tasks that require processing a large set or continuous stream of information. A common approach to reducing the size of data is through lossless or lossy compression. Yet, in some cases it may not be strictly necessary to perfectly recover every detail from the original data, as long as a requisite level of semantic precision or intent is conveyed. This paper presents three contributions to research on LLMs. First, we present the results from experiments exploring the viability of approximate compression using LLMs, focusing specifically on GPT-3.5 and GPT-4 via ChatGPT interfaces. Second, we investigate and quantify the capability of LLMs to compress text and code, as well as to recall and manipulate compressed representations of prompts. Third, we present two novel metrics -- Exact Reconstructive Effectiveness (ERE) and Semantic Reconstruction Effectiveness (SRE) -- that quantify the level of preserved intent between text compressed and decompressed by the LLMs we studied. Our initial results indicate that GPT-4 can effectively compress and reconstruct text while preserving the semantic essence of the original text, providing a path to leverage ~5× more tokens than present limits allow.