Elamur Research⌕
☰
home / authors / Tao Zhang

Tao Zhang

Tao Zhang has 8 papers in our library, in 2024–2026, most of them on Language models & pretraining, Reasoning, Speech & spoken language; often with Mingan Lin, Tianpeng Li, Guosheng Dong.

8 papers in our library · 2024–2026

OpenAlex · ORCID · h-index 49 · 526 works · 9,665 citations OpenAlex, 2026-10-14

📄 8 works in our library · 🔗 Cited 130 times in our library · ⭐ Paper Stars ranking: #6389

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 · China2014–2026 · 8 paperscurrent🏥Shanghai First People's HospitalShanghai · China2018–2026 · 6 paperscurrent🏥Affiliated Hospital of Xuzhou Medical CollegeXuzhou · China2026 · 1 papercurrent🎓Chongqing Normal UniversityChongqing · China2026 · 1 papercurrent🎓Huazhong University of Science and TechnologyWuhan · China2009–2025 · 45 papers🎓Chinese Academy of Medical Sciences & Peking Union Medical CollegeBeijing · China2014–2025 · 38 papers🏥Wuhan Union HospitalWuhan · China2009–2025 · 38 papers🏥First Hospital of Lanzhou UniversityLanzhou · China2014–2025 · 37 papersEmployment🏥Cancer Hospital of Chinese Academy of Medical SciencesBeijing · China2014–2025 · 29 papers🎓Lanzhou UniversityLanzhou · China2012–2025 · 21 papers

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 207 more institutions on single bylines are not listed.

 Fields, topics and concepts

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

📚 Shelves 8The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 8 worksComputer Science · 8 worksArtificial Intelligence · 8 worksLanguage and Linguistics · 8 worksComputer Vision and Pattern Recognition · 3 worksPsychology · 1 workSignal Processing · 1 workExperimental and Cognitive Psychology · 1 work

🧭 Topics 8Research subjects and languages the work is about.Language models & pretraining · 3 worksSpeech & spoken language · 1 workReasoning · 2 worksInformation retrieval & search · 1 workEvaluation & benchmarks · 1 workVision & language · 1 workEfficient models & compression · 1 workBias, fairness & safety · 1 work

 Works by year

2020s

20262 papers20254 papers20242 papers

 Papers

The most cited first (OpenAlex and our library).

  1. RAG-Star: Enhancing Deliberative Reasoning with Retrieval Augmented Verification and Refinement (2025 · 31 citations)
    Large language models often fail on multi-hop questions that require chaining several facts. Tree-search methods such as MCTS help with deliberative reasoning but rely only on the model's internal knowledge. Standard RAG feeds documents straight into generation, which can cause conflicts between retrieved and…
  2. CFBench: A Comprehensive Constraints-Following Benchmark for LLMs (2025 · 30 citations)
    CFBench tackles a gap in how LLMs are tested on following instructions that carry many explicit constraints.
  3. Baichuan-Omni-1.5 Technical Report (2025 · 28 citations)
    Open-source multimodal LLMs mostly pair vision with text. Voice support usually relies on separate ASR and TTS modules, which add latency. End-to-end omni models such as VITA-1.5 and Mini-Omni2 often lose quality when modalities compete, especially on text-only tasks. Baichuan-Omni-1.5 is a single 7B-scale model that…
  4. Baichuan-Omni Technical Report (2024 · 18 citations)
    GPT-4o showed what a single model handling several input types can do, but the open-source community had no comparable system.
  5. Baichuan-Audio: A Unified Framework for End-to-End Speech Interaction (2025 · 16 citations)
    Cascaded voice assistants chain ASR, an LLM and TTS. This adds latency, compounds errors and discards paralinguistic cues. End-to-end alternatives such as Moshi and GLM-4-Voice tend to lose reasoning ability once speech is added. Baichuan-Audio is a 7B end-to-end audio LLM for real-time bilingual (Chinese/English)…
  6. MixKVQ: Query-Aware Mixed-Precision KV Cache Quantization for Long-Context Reasoning (2026 · 5 citations)
    Reasoning-oriented LLMs produce very long Chain-of-Thought outputs, so their KV cache consumes large amounts of GPU memory and bandwidth.
  7. Baichuan Alignment Technical Report (2024 · 5 citations)
    Alignment methods for large language models are usually kept proprietary or described only in fragments. This technical report from Baichuan Inc. and Peking University sets out the full alignment pipeline behind the Baichuan models. The pipeline has three stages. A Prompt Augmentation System (PAS) automatically…
  8. From Narrow Unlearning to Emergent Misalignment in LLMs (2026)
    Earlier work found that fine-tuning an aligned LLM on a narrow task, such as writing insecure code, can cause emergent misalignment (EMA): harmful behaviour that spreads to unrelated domains.

 Co-authors here

12 of 39 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 130 times (of the 149,439 papers whose bibliographies we read); 39 citing papers with a page here →.

Updated

↑ back to top

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