# Token Sugar: Making Source Code Sweeter for LLMs through Token-Efficient Shorthand

arXiv: https://arxiv.org/abs/2512.08266 · HTML: https://mlsanity.com/paper/2512.08266/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Dec 2025.

## Built on by (most cited first, 4 of 4)

- [Seeing is Coding: On the Effectiveness of Vision Language Models in Code Understanding](https://mlsanity.com/paper/2602.01785/) (2026)
- [The Best Programming Language for Tokenmaxxing: An Investigation of Coding Agent Behavior Across Programming Languages](https://mlsanity.com/paper/2607.22807/) (2026)
- [CodeShrink: Adaptive Visual Compression for Efficient Multimodal Code Understanding](https://mlsanity.com/paper/2607.29637/) (2026)
- [Mind the Gaps: From Failure Attribution to Closed-Form Repair of Code Language Models](https://mlsanity.com/paper/2610.05277/) (2026)

## Abstract (arXiv)

Large language models (LLMs) have shown exceptional performance in code generation and understanding tasks, yet their high computational costs hinder broader adoption. One important factor is the inherent verbosity of programming languages, such as unnecessary formatting elements and lengthy boilerplate code. This leads to inflated token counts in both input and generated outputs, which increases inference costs and slows down the generation process. Prior work improves this through simplifying programming language grammar, reducing token usage across both code understanding and generation tasks. However, it is confined to syntactic transformations, leaving significant opportunities for token reduction unrealized at the semantic level. In this work, we propose Token Sugar, a concept that replaces frequent and verbose code patterns with reversible, token-efficient shorthand in the source code. To realize this concept in practice, we designed a systematic solution that mines high-frequency, token-heavy patterns from a code corpus, maps each to a unique shorthand, and integrates them into LLM pretraining via code transformation. With this solution, we obtain 799 (code pattern, shorthand) pairs, which can reduce up to 15.1% token count in the source code and is complementary to existing syntax-focused methods. We further trained three widely used LLMs on Token Sugar-augmented data. Experimental results show that these models not only achieve significant token savings (up to 11.2% reduction) during generation but also maintain near-identical Pass@1 scores compared to baselines trained on unprocessed code.