Elamur Research⌕
☰
home / authors / Xiaofeng Zhang

Xiaofeng Zhang

Xiaofeng Zhang has 6 papers in our library, in 2023–2026, most of them on Reasoning, Language models & pretraining; often with Xiaosong Yuan, Chen Shen, Yihao Quan.

6 papers in our library · 2023–2026

OpenAlex · h-index 1 · 19 works · 3 citations OpenAlex, 2026-10-14

📄 6 works in our library · 🔗 Cited 128 times in our library · 1 more in our library without a page yet · ⭐ Paper Stars ranking: #6159

The profile is the library's match of the name; two people with one name are not always told apart.

Elsewhere: Google Scholar · Semantic Scholar · dblp · ACL Anthology

 Affiliations

🎓Shanghai Jiao Tong UniversityShanghai · China2026 · 11 paperscurrent🎓Harbin Institute of TechnologyHarbin · China2026 · 3 paperscurrent

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine.

 Fields, topics and concepts

Of its works in our library, with the number of works on each.

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 6 worksArts and Humanities · 4 worksArtificial Intelligence · 5 worksLanguage and Linguistics · 4 worksComputer Vision and Pattern Recognition · 4 works

🧭 Topics 2Research subjects and languages the work is about.Reasoning · 2 worksLanguage models & pretraining · 1 work

 Works by year

2020s

20263 papers20252 papers20231 paper

 Papers

The most cited first (OpenAlex and our library).

  1. Transformer-Patcher: One Mistake worth One Neuron (2023 · 123 citations)
    Deployed language models keep producing new errors, yet most Model Editing (ME) research fixes only a single mistake at a time.
  2. Shallow Focus, Deep Fixes: Enhancing Shallow Layers Vision Attention Sinks to Alleviate Hallucination in LVLMs (2025 · 14 citations)
    The paper investigates why large vision-language models (LVLMs) hallucinate, focusing on attention over image tokens, which make up most of the input.
  3. From Redundancy to Relevance: Information Flow in LVLMs Across Reasoning Tasks (2025 · 12 citations)
    Large vision-language models (LVLMs) turn each image into hundreds of tokens before passing it to an LLM. How those visual tokens actually shape the generated answer has been unclear. The authors study this "information flow", meaning how strongly system, image and user-prompt tokens influence answer tokens, layer by…
  4. ART: Attention Replacement Technique to Improve Factuality in LLMs (2026 · 1 citation)
    This paper links hallucination in LLMs to the way attention is distributed in the earliest layers. Inspecting attention maps of shallow layers (e.g., Layer 0 of Qwen2.5-7B-Instruct on a GSM8K sample), the authors group heads into three types. Uniform heads spread weight evenly over the context, local heads look…
  5. Diffusion-CAM: Faithful Visual Explanations for dMLLMs (2026 · 1 citation)
    Diffusion-based multimodal LLMs (dMLLMs) such as LaViDa, LLaDA-V, MMaDA and Dream-VL produce answers by denoising all masked response tokens in parallel.
  6. Reasoning Fails Where Step Flow Breaks (2026 · 1 citation)
    Large reasoning models (LRMs) produce long question–thinking–summary traces, but their mistakes are hard to trace back to specific parts of the reasoning.

All works

7 works of theirs in our library, the most cited first; 6 with a page here, the others link their original until we summarise them.

  1. Transformer-Patcher: One Mistake worth One NeuronarXiv:2301.09785Zeyu Huang, Yikang Shen, Xiaofeng Zhang, et al.2023 · 123 citations
  2. Shallow Focus, Deep Fixes: Enhancing Shallow Layers Vision Attention Sinks to Alleviate Hallucination in LVLMsConference on Empirical Methods in Natural Language Processing (EMNLP)Xiaofeng Zhang, Yihao Quan, Chen Shen, et al.2025 · 14 citations
  3. From Redundancy to Relevance: Information Flow in LVLMs Across Reasoning TasksConference of the North American Chapter of the Association for Computational Linguistics (NAACL)Xiaofeng Zhang, Yihao Quan, Chen Shen, et al.2025 · 12 citations
  4. ART: Attention Replacement Technique to Improve Factuality in LLMsAnnual Meeting of the Association for Computational Linguistics (ACL)Ziqin Luo, Yihao Quan, Xiaofeng Zhang, et al.2026 · 1 citation
  5. Diffusion-CAM: Faithful Visual Explanations for dMLLMsAnnual Meeting of the Association for Computational Linguistics (ACL)Haomin Zuo, Yidi Li, Luoxiao Yang, et al.2026 · 1 citation
  6. Reasoning Fails Where Step Flow BreaksAnnual Meeting of the Association for Computational Linguistics (ACL)Xiaoyu Xu, Yulan Pan, Xiaosong Yuan, et al.2026 · 1 citation
  7. Fixing Semantic Blind Spots in Anchor Tokens of dMLLMs ↗Findings of the Association for Computational Linguistics: ACL 2026 · no summary yetRuixuan Xu et al.2026

 Co-authors here

12 co-authors with a page, the most papers together first.

Venues

Where their works here were published, the most works first, with the editions of those that have a page.

 Cited by

Papers of our library cite their works 128 times (of the 149,439 papers whose bibliographies we read); 24 citing papers with a page here →.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/authors/xiaofeng-zhang/