Pith. sign in

REVIEW 3 cited by

Efficient and Asymptotically Unbiased Constrained Decoding for Large Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2504.09135 v1 pith:AXQNGKEP submitted 2025-04-12 cs.CL

Efficient and Asymptotically Unbiased Constrained Decoding for Large Language Models

classification cs.CL
keywords constraineddecodingapplicationsdynamicexistinggenerationgpu-basedimportance
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

In real-world applications of large language models, outputs are often required to be confined: selecting items from predefined product or document sets, generating phrases that comply with safety standards, or conforming to specialized formatting styles. To control the generation, constrained decoding has been widely adopted. However, existing prefix-tree-based constrained decoding is inefficient under GPU-based model inference paradigms, and it introduces unintended biases into the output distribution. This paper introduces Dynamic Importance Sampling for Constrained Decoding (DISC) with GPU-based Parallel Prefix-Verification (PPV), a novel algorithm that leverages dynamic importance sampling to achieve theoretically guaranteed asymptotic unbiasedness and overcomes the inefficiency of prefix-tree. Extensive experiments demonstrate the superiority of our method over existing methods in both efficiency and output quality. These results highlight the potential of our methods to improve constrained generation in applications where adherence to specific constraints is essential.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Schema Key Wording as an Instruction Channel in Structured Generation under Constrained Decoding

    cs.CL 2026-04 unverdicted novelty 7.0

    Schema-key wording functions as an implicit instruction channel under constrained decoding, with experiments showing that rephrasing only the keys can substantially change accuracy on math benchmarks while prompt, mod...

  2. The Format Tax

    cs.CL 2026-04 conditional novelty 6.0

    Structured-output instructions alone impose a large accuracy tax on open-weight LLMs; decoupling freeform reasoning from formatting recovers most of it, while recent closed models largely avoid the tax.

  3. Vectorizing the Trie: Efficient Constrained Decoding for LLM-based Generative Retrieval on Accelerators

    cs.IR 2026-02 conditional novelty 6.0

    Constrained decoding for generative retrieval can be made accelerator-friendly by flattening the trie of valid items into a CSR sparse matrix and doing branch-free vectorized lookups.