PeopleofSuperintelligence

Researcher

Donovan Ong

Donovan Ong is an AI researcher whose work spans machine alignment, conversational emotion recognition, text-to-SQL, and multimodal language models. He coauthored “Would you Rather?” in 2020, studied discourse roles for conversational emotion recognition at A*STAR, and coauthored a text-to-SQL comparison in December 2023. In 2024, he coauthored Reka’s report on Core, Flash, and Edge and the Vibe-Eval benchmark for multimodal chat. As of October 2026, he is a Senior Research Engineer at A*STAR’s Institute for Infocomm Research, where he leads industry collaboration projects and researches deep-learning-based natural language processing.

Works
7
Citations
57
h-index
3

Publication numbers from OpenAlex, as of 19 October 2026.

Timeline

  1. 2020

    Cultural values benchmark

    In 2020, he coauthors “Would you Rather?”, a benchmark that learns alignments between machine and human preferences through choices between sentences describing social and cultural scenarios.

  2. 2022

    Discourse-role emotion recognition

    In 2022, he coauthors a paper that uses a variational autoencoder to model each utterance’s latent discourse role for emotion recognition, beating reported results on three public conversation datasets.

  3. March 2023

    An I2R conversation model

    In March 2023, he and colleagues in A*STAR’s I2R Natural Language Processing Group build a model that uses a variational autoencoder to label discourse roles and a recurrent neural network to model their sequence.

  4. December 2023

    LLMs for text-to-SQL

    In December 2023, he and six coauthors compare six large language models on text-to-SQL parsing across nine benchmark datasets and five prompting strategies.

  5. April 2024

    Reka multimodal models

    In April 2024, he coauthors Reka’s report introducing Core, Flash, and Edge as multimodal language models that process and reason with text, images, video, and audio.

  6. May 2024

    Vibe-Eval benchmark

    In May 2024, he coauthors Vibe-Eval, an open benchmark for multimodal chat with 269 visual-understanding prompts, including 100 hard prompts.

  7. As of October 2026

    Senior Research Engineer at I2R

    As of October 2026, he is a Senior Research Engineer at A*STAR’s Institute for Infocomm Research (I2R), leads industry collaboration projects, and researches deep-learning-based natural language processing with an international collaboration.

Career

  1. Since 2022Agency for Science, Technology and ResearchSenior Research Engineer
  2. 2022–2024Science and Engineering Research Council

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. Is Discourse Role Important for Emotion Recognition in Conversation?2022 · Proceedings of the AAAI Conference on Artificial Intelligence · 32 citationsWith Haifeng Wang
  2. Would you Rather? A New Benchmark for Learning Machine Alignment with Cultural Values and Social Preferences2020 · Annual Meeting of the Association for Computational Linguistics (ACL) · 12 citationsWith Nancy Chen
  3. Battle of the Large Language Models: Dolly vs LLaMA vs Vicuna vs Guanaco vs Bard vs ChatGPT - A Text-to-SQL Parsing Comparison2023 · 9 citations
  4. Reka Core, Flash, and Edge: A Series of Powerful Multimodal Language Models2024 · 3 citationsWith Aitor Ormazabal, Che Zheng, Cyprien de Masson d’Autume and Dani Yogatama
  5. Vibe-Eval: A hard evaluation suite for measuring progress of multimodal language models2024 · 1 citationsWith Aitor Ormazabal, Che Zheng, Cyprien de Masson d’Autume and Dani Yogatama

From OpenAlex, as of 19 October 2026.

  1. The company share of AI researchers’ top papers rose from 2% to 32%Includes Agency for Science, Technology and Research. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 2% in 1975–1979 to 32% in 2020–2024.
  2. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  3. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  4. 47% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Donovan Ong yet. Send us a free-licensed photo.

  1. Would you Rather? A New Benchmark for Learning Machine Alignment with Cultural Values and Social Preferences aclanthology.org. Checked 19 October 2026.
  2. Is Discourse Role Important for Emotion Recognition in Conversation? ojs.aaai.org. Checked 19 October 2026.
  3. Artificial intelligence that knows how you feel research.a-star.edu.sg. Checked 19 October 2026.
  4. Battle of the Large Language Models: Dolly vs LLaMA vs Vicuna vs Guanaco vs Bard vs ChatGPT - A Text-to-SQL Parsing Comparison aclanthology.org. Checked 19 October 2026.
  5. Reka Core, Flash, and Edge: A Series of Powerful Multimodal Language Models arxiv.org. Checked 19 October 2026.
  6. Vibe-Eval: A hard evaluation suite for measuring progress of multimodal language models arxiv.org. Checked 19 October 2026.
  7. Donovan Ong - A*STAR Research research.a-star.edu.sg. Checked 19 October 2026.