Elamur Research⌕
☰
home / authors / Han Wang

Han Wang

Han Wang has 6 papers in our library, in 2018–2026, most of them on Reasoning, Language models & pretraining, Speech & spoken language; often with Changhao Pan, Chenyuhao Wen, Haiyang Yu.

6 papers in our library · 2018–2026

OpenAlex · ORCID · h-index 30 · 187 works · 3,225 citations OpenAlex, 2026-10-14

📄 6 works in our library · 🔗 Cited 67 times in our library · ⭐ Paper Stars ranking: #7015

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

🎓Guangxi UniversityNanning · China2018–2026 · 41 paperscurrent🎓University of North Carolina at Chapel HillChapel Hill · United States2018–2026 · 34 paperscurrent🎓Hangzhou City UniversityHangzhou2023–2026 · 6 paperscurrent🎓University of JyväskyläJyväskylä · Finland2024–2026 · 3 paperscurrent🎓Central South UniversityChangsha · China2025–2026 · 2 paperscurrent🎓Changchun University of Chinese MedicineChangchun · China2020–2026 · 2 paperscurrent🎓North China University of Science and TechnologyTangshan · China2023–2026 · 2 paperscurrent🎓Northeast Agricultural UniversityHarbin · China2017–2026 · 2 paperscurrent🎓Northeast Normal UniversityChangchun · China2011–2026 · 2 paperscurrent🎓University of MalayaKuala Lumpur · Malaysia2025–2026 · 2 paperscurrent

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 69 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 5The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 6 worksComputer Science · 6 worksArtificial Intelligence · 6 worksLanguage and Linguistics · 6 worksComputer Vision and Pattern Recognition · 2 works

🧭 Topics 7Research subjects and languages the work is about.Reasoning · 3 worksLanguage models & pretraining · 2 worksBias, fairness & safety · 1 workEfficient models & compression · 1 workEnglish · 1 workEvaluation & benchmarks · 1 workSpeech & spoken language · 1 work

 Works by year

2020s

20262 papers20252 papers20241 paper

2010s

20181 paper

 Papers

The most cited first (OpenAlex and our library).

  1. Self-Distillation Bridges Distribution Gap in Language Model Fine-Tuning (2024 · 63 citations)
    When an instruction-tuned LLM is further fine-tuned for a narrow downstream task, it tends to lose general abilities such as instruction following, helpfulness and safety alignment.
  2. Prolonged Reasoning Is Not All You Need: Certainty-Based Adaptive Routing for Efficient LLM/MLLM Reasoning (2025 · 11 citations)
    Reasoning-tuned LLMs and MLLMs tend to produce long chain-of-thought output for every query. This costs many tokens and can even hurt accuracy on easy, fact-based questions. A pilot with Qwen2.5-0.5B on text-rich VQA and KIE data showed that direct answers were both more accurate and far cheaper than reasoning-based…
  3. AlphaOne: Reasoning Models Thinking Slow and Fast at Test Time (2025 · 10 citations)
    Models like OpenAI o1 and DeepSeek-R1, known as large reasoning models (LRMs), frequently overthink or underthink at inference time.
  4. English Out-of-Vocabulary Lexical Evaluation Task (2018 · 3 citations)
    The paper proposes a benchmark for words a model has never seen: out-of-vocabulary (OOV) items that occur only in test material, so their meaning must come from surrounding text.
  5. Comprehensive Benchmarking of Long-Form Speech Generation in Diverse Scenarios (2026)
    Long-form text-to-speech has shifted from single sentences to paragraphs and multi-speaker dialogue, but evaluation has not kept pace.
  6. Stabilizing Efficient Reasoning with Step-Level Advantage Selection (2026)
    Methods that make LLM reasoning shorter usually add length penalties or pruning. They also tend to post-train in a much smaller context window than the base model used. The authors separate these two factors. They post-train DeepScaleR-1.5B-Preview with plain GRPO in a 4K window, using only a correctness reward. This…

 Co-authors here

12 of 23 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 67 times (of the 149,439 papers whose bibliographies we read); 14 citing papers with a page here →.

Updated

↑ back to top

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