# Compressed code: the hidden effects of quantization and distillation on programming tokens

arXiv: https://arxiv.org/abs/2601.02563 · HTML: https://mlsanity.com/paper/2601.02563/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 18 references that are in the map.
- First posted on arXiv 5 Jan 2026.

## Builds on (most cited first, 14 of 14)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference](https://mlsanity.com/paper/2403.04132/) (2024)
- [DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model](https://mlsanity.com/paper/2405.04434/) (2024)
- [BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions](https://mlsanity.com/paper/2406.15877/) (2024)
- [Byte Pair Encoding is Suboptimal for Language Model Pretraining](https://mlsanity.com/paper/2004.03720/) (2020)
- [Byte Latent Transformer: Patches Scale Better Than Tokens](https://mlsanity.com/paper/2412.09871/) (2024)
- [Model Collapse Demystified: The Case of Regression](https://mlsanity.com/paper/2402.07712/) (2024)
- [Getting the most out of your tokenizer for pre-training and domain adaptation](https://mlsanity.com/paper/2402.01035/) (2024)
- [Low-Bit Quantization Favors Undertrained LLMs: Scaling Laws for Quantized LLMs with 100T Training Tokens](https://mlsanity.com/paper/2411.17691/) (2024)
- [Smaller = Weaker? Benchmarking Robustness of Quantized LLMs in Code Generation](https://mlsanity.com/paper/2506.22776/) (2025)
- [Qtok: A Comprehensive Framework for Evaluating Multilingual Tokenizer Quality in Large Language Models](https://mlsanity.com/paper/2410.12989/) (2024)

## Built on by (most cited first, 1 of 1)

- [Cultural Awareness is Represented but Not Decoded: Tracing Mythological Knowledge across 18 Open-Source LLMs](https://mlsanity.com/paper/2608.02486/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) have demonstrated exceptional code generation capabilities, yet their token-level mechanisms remain underexplored, particularly in compressed models. Through systematic analysis of programming language token representations, we characterize how programming languages are encoded in LLM tokenizers by analyzing their vocabulary distribution and keyword coverage patterns. We introduce a novel cold-start probability analysis method that provides insights into model behavior without requiring explicit prompts. Additionally, we present a comprehensive evaluation of how different model optimization techniques - including quantization, distillation, model scaling, and task-specific fine-tuning - affect token-level representations and code generation quality. Our experiments, supported by comprehensive probability distribution analysis and evaluation metrics, reveal critical insights into token-level behavior and provide empirically-validated guidelines for maintaining code generation quality under various optimization constraints. These findings advance both theoretical understanding of LLM code generation and practical implementation of optimized models in production environments.