Yiming Cui

Competition Awards
  • new A-OKVQA Challenge
    • [Nov, 2023] VL-Reasoner got First Place on A-OKVQA, a knowledge-based visual QA challenge.
  • DROP Challenge
    • [Jun, 2023] D-Reasoner got First Place on DROP, a discrete reasoning QA benchmark.
  • QASC Challenge
    • [Oct, 2022] X-Reasoner++ got First Place on QASC, becoming the first model system that surpasses human performance!Snapshot
  • OpenBookQA Challenge
    • [July, 2022] X-Reasoner got First Place on OpenBookQA, becoming the first single-model system that surpasses human performance!Snapshot
  • SemEval 2022 (16th International Workshop on Semantic Evaluation)
    •
  • VCR (Visual Commonsense Reasoning)
    • [Jan, 2022] VL-RoBERTa got First Place on VCR with Q2AR score of 72.8! Snapshot
  • XTREME (Cross-Lingual Transfer Evaluation of Multilingual Encoders)
    • [Oct, 2021] CoFe got First Place on XTREME with an average score of 84.1! Snapshot
  • GLUE (General Language Understanding Evaluation)
    • [Aug, 2020] MacALBERT + DKM got First Place on GLUE with an average score of 90.7! Snapshot
  • HotpotQA (Diverse, Explainable Multi-hop Question Answering)
    • [May, 2020] Step-by-Step Retriever got First Place on 'Fullwiki Setting' track. Snapshot PDF
    • [Oct, 2019] C2F Reader got First Place on 'Distractor Setting' track. Snapshot PDF
  • SQuAD (Stanford Question Answering Dataset) 2.0
    • [Mar, 2019] BERT + DAE + AoA got First Place, the first system outperforms human performance. Snapshot PDF
    • [Nov, 2018] AoA + DA + BERT got First Place, the first system that F1 exceeds 85%. Snapshot PDF
  • CoQA CoQA (Conversational Question Answering Challenge)
    • [Dec, 2018] D-AoA + BERT got First Place, the first system that overall score exceeds 80%. Snapshot PDF
  • QuAC QuAC (Question Answering in Context)
    •
  • SQuAD (Stanford Question Answering Dataset) 1.1
    • [Jan, 2018] Hybrid AoA Reader got First Place, the first system that F1 exceeds 89%! Snapshot
    • [Oct, 2017] Interactive AoA Reader+ got First Place, the first system that EM exceeds 79%. Snapshot
    • [Jul, 2017] Interactive AoA Reader got First Place, the first system that F1 exceed1 85%. Snapshot
  • SemEval 2018 (12th International Workshop on Semantic Evaluation: Task 11)
    • [Feb, 2018] Our system got First Place in Task 11: Machine Comprehension using Commonsense Knowledge.
  • NIST OpenMT15 Evaluation
    •
  • IWSLT 2014 (11th International Workshop on Spoken Language Translation)
    • [Dec, 2014] First Place in Chinese-to-English and English-to-Chinese TED translation task (participated as USTC). Result (PDF)
  • IWSLT 2012 (9th International Workshop on Spoken Language Translation)
    • [Dec, 2012] First Place in Chinese-to-English Olympic translation task (participated as HIT-LTRC). Result (PDF)

    Other Awards
  • [2023] One of "World top 2% scientists", Stanford University & Elsevier
    •
  • [2021] Outstanding Author Award, Publishing House of Electronics Industry
    •
  • [2021] Featured Researcher of TPU Research Cloud (TRC) Program, Google
  • [2018] The First HuaXia Innovation Award (10 out of 10,000+), iFLYTEK LTD, China
    •
  • [2017] Outstanding Team Award (as leader), iFLYTEK Research, China
    •
  • [2016] Outstanding Employee Award, iFLYTEK Research, China
    •
  • [2015] Outstanding Employee Award, iFLYTEK Research, China
    •
  • [2014] Outstanding New Employee Award, iFLYTEK Research, China
    •
  • [2014] Outstanding Graduates Award (Gold Medal), Harbin Institute of Technology, China
    •
  • [2014] Province-Level Triple-A Student Award, Heilongjiang Province, China
    •
  • [2013] Special Grade Scholarship, Harbin Institute of Technology, China
    •
  • [2013] GuangHua Scholarship, Harbin Institute of Technology, China
    •
  • [2012] Second Grade Scholarship, Harbin Institute of Technology, China
    •