paper-with-me

Papers

Information-Guided Frontier Decoding: Contextual Utility-Driven Commitment in dMLLMs

2026-08-27 · Xingyou Fang, Jingxing Zhong, Xiaosong Yuan, Xiaofeng Zhang arxiv

Decoding quality in diffusion multimodal language models (dMLLMs) depends heavily on the order in which masked tokens are committed. Existing confidence-based strategies prioritize locally easy tokens, but confidence does not necessarily reflect contextual usefulness. As a result, structurally easy tokens such as punctuation may be committed before informative semantic anchors, weakening context propagation and increasing error accumulation. We propose Information-Guided Frontier Decoding (IGFD), a training-free decoding strategy that ranks candidates using token confidence, neighborhood uncertainty, and structural commitment risk. IGFD encourages early commitment of reliable semantic anchors while delaying fragile structural tokens, improving contextual support during decoding. A dynamic candidate frontier further constrains token selection to locally expandable regions under the same decoding budget. The method requires no additional training, auxiliary models, or extra forward passes. Experiments across multimodal understanding, reasoning, grounding, and hallucination benchmarks show that IGFD consistently outperforms existing decoding strategies across the majority of benchmarks and diffusion MLLM backbones under identical decoding budgets.

📄 PDF Abstract BibTeX arXiv:2608.26641

Code (2)

Tavish9/awesome-daily-AI-arxiv ★ 113
arxivsub/arXivSub_daily_arxiv ★ 4

Similar Papers 제목 키워드 기반

PromptGraph: Graph-Guided Prompt Sanitization for Balancing Privacy and Utility in LLM Inference

2026-07-12 · Chen Gu, Hui Wan, Donghui Hu, Hui Wang 외 arxiv

Large Language Model (LLM) services introduce a fundamental privacy challenge. Sensitive information may be inferred not only from explicit identifiers, such as names or phone numbers, but also from contextual associatio…

Anchored Decoding: Provably Reducing Copyright Risk for Any Language Model

2026-02-06 · Jacqueline He, Jonathan Hayase, Wen-tau Yih, Sewoong Oh 외 arxiv

Language models (LMs) tend to memorize portions of their training data and emit verbatim spans. When the underlying sources are sensitive or copyright-protected, such reproduction raises issues of consent and compensatio…

Language Model Decoding as Likelihood-Utility Alignment

2022-10-13 · Martin Josifoski, Maxime Peyrard, Frano Rajic, Jiheng Wei 외

A critical component of a successful language generation pipeline is the decoding algorithm. However, the general principles that should guide the choice of a decoding algorithm remain unclear. Previous works only compar…

Language ModelingLanguage ModellingmodelText Generation

Seeing is Believing: Mitigating Hallucination in Large Vision-Language Models via CLIP-Guided Decoding

2024-02-23 · Ailin Deng, Zhirui Chen, Bryan Hooi

Large Vision-Language Models (LVLMs) are susceptible to object hallucinations, an issue in which their generated text contains non-existent objects, greatly limiting their reliability and practicality. Current approaches…

HallucinationObjectObject HallucinationSentence+2

Dynamic Attention-Guided Context Decoding for Mitigating Context Faithfulness Hallucinations in Large Language Models

2025-01-02 · Yanwen Huang, Yong Zhang, Ning Cheng, Zhitao Li 외

Large language models (LLMs) often suffer from context faithfulness hallucinations, where outputs deviate from retrieved information due to insufficient context utilization and high output uncertainty. Our uncertainty ev…

Computational Efficiency