Constrained Decoding of Diffusion LLMs with Context-Free Grammars

• URL: http://arxiv.org/abs/2508.10111v1
• Date: Wed, 13 Aug 2025 18:09:09 GMT
• Title: Constrained Decoding of Diffusion LLMs with Context-Free Grammars
• Authors: Niels Mündler, Jasper Dekoninck, Martin Vechev,
• Abstract summary: Large language models (LLMs) have shown promising performance across diverse domains.<n>This paper presents the first constrained decoding method for diffusion models.<n>We show that our method achieves near-perfect syntactic correctness while consistently preserving or improving functional correctness.
• Score: 1.0923877073891446
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) have shown promising performance across diverse domains. Many practical applications of LLMs, such as code completion and structured data extraction, require adherence to syntactic constraints specified by a formal language. Yet, due to their probabilistic nature, LLM output is not guaranteed to adhere to such formal languages. Prior work has proposed constrained decoding as a means to restrict LLM generation to particular formal languages. However, existing works are not applicable to the emerging paradigm of diffusion LLMs, when used in practical scenarios such as the generation of formally correct C++ or JSON output. In this paper we address this challenge and present the first constrained decoding method for diffusion models, one that can handle formal languages captured by context-free grammars. We begin by reducing constrained decoding to the more general additive infilling problem, which asks whether a partial output can be completed to a valid word in the target language. This problem also naturally subsumes the previously unaddressed multi-region infilling constrained decoding. We then reduce this problem to the task of deciding whether the intersection of the target language and a regular language is empty and present an efficient algorithm to solve it for context-free languages. Empirical results on various applications, such as C++ code infilling and structured data extraction in JSON, demonstrate that our method achieves near-perfect syntactic correctness while consistently preserving or improving functional correctness. Importantly, our efficiency optimizations ensure that the computational overhead remains practical.

Related papers

• Lookahead-then-Verify: Reliable Constrained Decoding for Diffusion LLMs under Context-Free Grammars [17.13122301190815]
  We present LAVE, a constrained decoding approach specifically designed for dLLMs.<n>Our approach leverages a key property of dLLMs, namely their ability to predict token distributions for all positions in parallel during each forward pass.<n>Extensive experiments across four widely used dLLMs and three representative benchmarks demonstrate that LAVE consistently outperforms existing baselines and achieves substantial improvements in syntactic correctness, while incurring negligible runtime overhead.
  arXiv  Detail & Related papers  (2026-01-31T08:58:15Z)
• Decoding in Latent Spaces for Efficient Inference in LLM-based Recommendation [75.72196852363116]
  Light Latent-space Decoding (L2D) is an effective and efficient latent-space decoding method.<n>L2D is more than 10x faster than language-space decoding while maintaining or enhancing performance.
  arXiv  Detail & Related papers  (2025-09-15T02:30:35Z)
• RELIC: Evaluating Compositional Instruction Following via Language Recognition [37.49115450182637]
  Large language models (LLMs) are increasingly expected to perform tasks based only on a specification of the task provided in context.<n>We introduce the Recognition of Languages In-Context (RELIC) framework to evaluate instruction following using language recognition.
  arXiv  Detail & Related papers  (2025-06-05T16:17:24Z)
• The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLMs [54.59207567677249]
  Large language models (LLMs) still struggle across tasks outside of high-resource languages.<n>In this work, we investigate cross-lingual transfer to lower-resource languages where task-specific post-training data is scarce.
  arXiv  Detail & Related papers  (2025-05-23T20:28:31Z)
• Type-Constrained Code Generation with Language Models [51.03439021895432]
  We introduce a type-constrained decoding approach that leverages type systems to guide code generation.<n>For this purpose, we develop novel prefix automata and a search over inhabitable types, forming a sound approach to enforce well-typedness on LLM-generated code.<n>Our approach reduces compilation errors by more than half and significantly increases functional correctness in code synthesis, translation, and repair tasks.
  arXiv  Detail & Related papers  (2025-04-12T15:03:00Z)
• Fast Controlled Generation from Language Models with Adaptive Weighted Rejection Sampling [90.86991492288487]
  evaluating constraint on every token can be prohibitively expensive.<n> LCD can distort the global distribution over strings, sampling tokens based only on local information.<n>We show that our approach is superior to state-of-the-art baselines.
  arXiv  Detail & Related papers  (2025-04-07T18:30:18Z)
• Automata-based constraints for language model decoding [9.137697105669142]
  Language models (LMs) are often expected to generate strings in some formal language. tuning requires significant resources, making it impractical for uncommon or task-specific formats. We solve these issues through the application of automata theory. Our system compiles constraints 7,000x faster, is provably correct, and can be extended in a modular fashion.
  arXiv  Detail & Related papers  (2024-07-11T00:25:01Z)
• Decoding at the Speed of Thought: Harnessing Parallel Decoding of Lexical Units for LLMs [57.27982780697922]
  Large language models have demonstrated exceptional capability in natural language understanding and generation. However, their generation speed is limited by the inherently sequential nature of their decoding process. This paper introduces Lexical Unit Decoding, a novel decoding methodology implemented in a data-driven manner.
  arXiv  Detail & Related papers  (2024-05-24T04:35:13Z)
• Guiding LLMs The Right Way: Fast, Non-Invasive Constrained Generation [7.687678490751105]
  We present a novel decoding algorithm, DOMINO, that can enforce constraints in a fully subword-aligned fashion, while leveraging pre-computation and speculative decoding to achieve virtually no overhead and in some cases even almost 2$times$ speedup over unconstrained decoding -- thereby outperforming existing approaches by a wide margin.
  arXiv  Detail & Related papers  (2024-02-07T13:36:02Z)
• The Ups and Downs of Large Language Model Inference with Vocabulary Trimming by Language Heuristics [74.99898531299148]
  This research examines vocabulary trimming (VT) inspired by restricting embedding entries to the language of interest to bolster time and memory efficiency. We apply two languages to trim the full vocabulary - Unicode-based script filtering and corpus-based selection - to different language families and sizes. It is found that VT reduces the memory usage of small models by nearly 50% and has an upper bound of 25% improvement in generation speed.
  arXiv  Detail & Related papers  (2023-11-16T09:35:50Z)
• Grammar-Constrained Decoding for Structured NLP Tasks without Finetuning [27.59524153097858]
  grammar-constrained decoding (GCD) can be used to control the generation of large language models (LMs) GCD can serve as a unified framework for structured NLP tasks in general. We show that grammar-constrained LMs substantially outperform unconstrained LMs or even beat task-specific finetuned models.
  arXiv  Detail & Related papers  (2023-05-23T11:54:37Z)
• LEVER: Learning to Verify Language-to-Code Generation with Execution [64.36459105535]
  We propose LEVER, a simple approach to improve language-to-code generation by learning to verify the generated programs with their execution results. Specifically, we train verifiers to determine whether a program sampled from the LLMs is correct or not based on the natural language input, the program itself and its execution results. LEVER consistently improves over the base code LLMs(4.6% to 10.9% with code-davinci) and achieves new state-of-the-art results on all of them.
  arXiv  Detail & Related papers  (2023-02-16T18:23:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.