Code generation

Read diffusion methods for program generation, completion or repair. DiffuCoder studies iterative denoising over code and training choices; identify the task contract before interpreting an evaluation score.

How to read and compare
  • Is the task completion, transformation, repair or infilling, and what context must stay unchanged?
  • Are syntax validity, execution or tests, and functional correctness reported separately?

A code-generation study does not guarantee correct or safe programs, or a general advantage over other generation methods.

A code link does not guarantee reproducibility. Check the original dataset, metric, protocol and conditions.

Primary references

These are reading anchors, not an exhaustive survey or ranking. A source reference does not imply it appears in the current selection.

Related topics

Selection methodology

2026-08-31T15:00:30+00:00 · Masked / discrete diffusion, Reasoning, Code generation · Source

CARVE: Verified Expansion for Variable-Length Generation in Diffusion Language Models

Masked diffusion language models predict tokens from a partially observed response canvas, enabling bidirectional conditioning and parallel token refinement. Yet standard masked-diffusion decoders use a rigid inference interface: the number of masked positions allocated to the answer is fixed before generation begins. Choosing this length is difficult. A short canvas can truncate reasoning or code, while a long canvas wastes computation and can perturb denoising. We introduce CARVE (Counterfactual-Aware Reveal with Verified Expansion), a training-free variable-length algorithm for masked diffu

Useful for: Not assessed · Limitation: Not assessed

2026-06-16T14:46:53+00:00 · Masked / discrete diffusion, Reasoning, Code generation · Source

VoidPadding: Let [VOID] Handle Padding in Masked Diffusion Language Models so that [EOS] Can Focus on Semantic Termination

MDLMs generate text by denoising a preallocated masked response canvas, making response-length modeling central to instruction tuning. Existing MDLMs often inherit the autoregressive convention of using repeated \texttt{[EOS]} tokens for padding during instruction tuning, giving \texttt{[EOS]} a dual role as both a semantic terminator and a padding token. We show that this dual role is a root cause of \texttt{[EOS]} overflow under large-block decoding. To decouple these roles, we propose VoidPadding, which introduces \texttt{[VOID]} for padding and reserves \texttt{[EOS]} for termination. Duri

Useful for: Not assessed · Limitation: Not assessed

2026-06-07T02:24:11+00:00 · Masked / discrete diffusion, Reasoning, Code generation · Source

AsyncLane: Decoupling Refinement from Advancement in Diffusion Language Model Decoding

Block-wise semi-autoregressive decoding is the standard inference paradigm for diffusion large language models (DLMs), but it imposes a strict dependency between blocks: the next block cannot begin until the current block is fully decoded or its denoising budget is exhausted. We observe that once a block exposes a reliable delimiter boundary or stable semantic prefix, continuation generation need not wait for every residual token to be resolved. We propose AsyncLane, a training-free decoding scheduler that decouples refinement from advancement. AsyncLane forks a generate lane at observed delim

Useful for: Not assessed · Limitation: Not assessed

2026-05-16T06:15:47+00:00 · Masked / discrete diffusion, Code generation · Source

Constrained Code Generation with Discrete Diffusion

Discrete diffusion models are a powerful, emerging paradigm for code generation. They construct programs through iterative refinement of partially corrupted token sequences and enable parallel token refinement. Importantly, this paradigm exposes a global program state at each denoising step, which provides a natural intervention point for enforcing program-level functionality and security constraints, guiding the generation before the final code is committed. Building on this observation, the paper introduces Constrained Diffusion for Code (CDC), a training-free neurosymbolic inference framewo

Useful for: Not assessed · Limitation: Not assessed