paper-with-me

홈 › Papers

Plan, Verify and Fill: A Structured Parallel Decoding Approach for Diffusion Language Models

2026-01-18 · Miao Li, Hanyang Jiang, Sikai Cheng, Hengyu Fu, Yuhang Cai, Baihe Huang, Tinghan Ye, Xuanzhou Chen, Pascal Van Hentenryck arxiv

Diffusion Language Models (DLMs) present a promising non-sequential paradigm for text generation, distinct from standard autoregressive (AR) approaches. However, current decoding strategies often adopt a reactive stance, underutilizing the global bidirectional context to dictate global trajectories. To address this, we propose Plan-Verify-Fill (PVF), a training-free paradigm that grounds planning via quantitative validation. PVF actively constructs a hierarchical skeleton by prioritizing high-leverage semantic anchors and employs a verification protocol to operationalize pragmatic structural stopping where further deliberation yields diminishing returns. Extensive evaluations on LLaDA-8B-Instruct and Dream-7B-Instruct demonstrate that PVF reduces the Number of Function Evaluations (NFE) by up to 65% compared to confidence-based parallel decoding across benchmark datasets, unlocking superior efficiency without compromising accuracy.

📄 PDF Abstract BibTeX arXiv:2601.12247

Code (0)

등록된 구현이 없습니다.

Tasks

Text Generation

Similar Papers 제목 키워드 기반

Reading Between the Dots: Decoding Hidden Computation across Filler Tokens

2026-07-03 · Kaley Brauer, Claudio Mayrink Verdun, Samuel Marks arxiv

Frontier LLMs can perform multi-step reasoning over content-free filler tokens like dots or counting sequences, producing correct answers with no visible chain-of-thought (CoT). This is a limit case for behavioral oversi…

Speculative Pipeline Decoding: Higher-Accruacy and Zero-Bubble Speculation via Pipeline Parallelism

2026-05-29 · Yijiong Yu, Huazheng Wang, Shuai Yuan, Ruilong Ren 외 arxiv

Speculative Decoding (SD) accelerates low-concurrency LLM inference by employing a draft-then-verify paradigm. However, mainstream methods typically rely on multi-token prediction, which introduces escalating prediction …

BlockPilot: Instance-Adaptive Policy Learning for Diffusion-based Speculative Decoding

2026-06-30 · Hao Zhang, Yiming Hu, Yong Wang, Mingqiao Mo 외 hf

Speculative decoding accelerates inference by using a lightweight draft model to generate candidate tokens in parallel, and are then verified by the target model, enabling lossless acceleration. Recently, diffusion-based…

Jupiter: Fast and Resource-Efficient Collaborative Inference of Generative LLMs on Edge Devices

2025-04-11 · Shengyuan Ye, Bei Ouyang, Liekang Zeng, Tianyi Qian 외

Generative large language models (LLMs) have garnered significant attention due to their exceptional capabilities in various AI tasks. Traditionally deployed in cloud datacenters, LLMs are now increasingly moving towards…

Collaborative InferenceEdge-computing

Copy-as-Decode: Grammar-Constrained Parallel Prefill for LLM Editing

2026-04-20 · Ziyang Liu arxiv

LLMs edit text and code by autoregressively regenerating the full output, even when most tokens appear verbatim in the input. We study Copy-as-Decode, a decoding-layer mechanism that recasts edit generation as structured…