# AdaDec: A Uncertainty-Guided Lookahead Decoding Framework for LLM-Based Code Generation

arXiv: https://arxiv.org/abs/2506.08980 · HTML: https://mlsanity.com/paper/2506.08980/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Jun 2025.

## Built on by (most cited first, 3 of 3)

- [EffiReasonTrans: RL-Optimized Reasoning for Code Translation](https://mlsanity.com/paper/2510.18863/) (2025)
- [Cut Your Losses! Learning to Prune Paths Early for Efficient Parallel Reasoning](https://mlsanity.com/paper/2604.16029/) (2026)
- [The Optimal Decoder is Tilted Temperature](https://mlsanity.com/paper/2610.15058/) (2026)

## Abstract (arXiv)

Code generation with large language models (LLMs) is highly sensitive to token selection during decoding, particularly at uncertain decision points that influence program logic. While standard strategies such as greedy decoding treat all tokens uniformly, they overlook code-specific uncertainty patterns, leading to suboptimal performance. This paper presents an empirical study revealing that many generation errors stem from token ranking mistakes at high-uncertainty steps, where the correct token is present but not top-ranked. Motivated by these findings, we propose AdaDec, a lookahead-based uncertainty-guided adaptive decoding framework that integrates a token-level pause-then-rerank mechanism driven by token uncertainty. AdaDec learns model-specific uncertainty thresholds and applies a lookahead-based reranking strategy when uncertainty is high. Experiments on HumanEval+, MBPP+, and DevEval benchmarks show that AdaDec improves Pass@1 accuracy by up to 20.9% in absolute terms over greedy decoding. More importantly, it consistently outperforms both competitive baselines like Beam Search and state-of-the-art adaptive decoding methods such as AdapT, while maintaining high efficiency through selective, uncertainty-triggered pausing. Our results highlight the promise of uncertainty-aware adaptive decoding for improving both the reliability and efficiency of LLM-based code generation.