PeopleofSuperintelligence

Researcher

Andrea Madotto

Andrea Madotto is an AI researcher whose work spans dialogue systems, multimodal language models and visual understanding. He studied computer science at Perugia University and Pisa University, completed a PhD at HKUST, and received two paper awards in 2019. He started at Meta FAIR in 2022 and researched multimodal language models. His later work includes AnyMAL, language models for materials generation, and PerceptionLM. As of October 2026, he is a Member of Technical Staff at Thinking Machines Lab.

Works
137
Citations
9,539
h-index
34
Badges

Publication numbers from OpenAlex, as of 18 October 2026.

Timeline

  1. 2019

    Dialogue systems and awards

    In 2019, he coauthors TRADE, a dialogue-state generator that transfers predictions to new domains. He receives the ACL outstanding paper award and the NeurIPS ConvAI workshop best paper award.

  2. 2020–2021

    A tutorial and a thesis

    He co-presents the NeurIPS 2020 tutorial Deeper Conversational AI. His education includes a computer-science BSc at Perugia University, an MSc in computer science at Pisa University, an MSc in AIS at HKBU, and a PhD in electronic and computer engineering at HKUST. His 2021 thesis, Taming the Beast: Learning to Control Neural Conversational Models, studies methods to control task-oriented and chit-chat dialogue systems.

  3. 2022–2023

    Research at Meta FAIR

    He starts at Meta FAIR as a research scientist in January 2022, where he works on multimodal language models and LLM-based agents that navigate user interfaces. In September 2023, he coauthors AnyMAL, a model that combines text, image, video, audio and motion inputs to generate text.

  4. 2024

    Language models for materials

    In 2024, he coauthors Fine-Tuned Language Models Generate Stable Inorganic Materials as Text, which fine-tunes language models on text-encoded atomistic data to generate stable materials.

  5. 2025

    Open visual understanding

    In 2025, he coauthors PerceptionLM, a vision-language project for image and video understanding. The project releases 2.8 million human-labeled instances, including video question-answer pairs and grounded captions.

  6. By April 2026

    Moving to Thinking Machines

    By April 2026, he has joined Thinking Machines Lab after working as a Meta FAIR research scientist on multimodal language models.

  7. As of October 2026

    Technical staff at Thinking Machines

    He is a Member of Technical Staff at Thinking Machines Lab.

Career

  1. Since 2026Thinking Machines LabMember of Technical Staff
  2. From 2022FAIR (Meta)Research Scientist
  3. 2022–2025Meta

Education

  1. University of PerugiaBSc in Computer Science
  2. University of PisaMSc in Computer Science
  3. Hong Kong Baptist UniversityMSc in AIS
  4. Hong Kong University of Science and TechnologyPhD

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

Most-cited works

  1. Survey of Hallucination in Natural Language Generation2022 · ACM Computing Surveys · 4,831 citationsWith Pascale Fung
  2. Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models2022 · 569 citationsWith Abhinav Rastogi, Abubakar Abid, Ashish Sabharwal, Chris Callison-Burch, Christopher D. Manning and Christopher Potts
  3. Plug and Play Language Models: A Simple Approach to Controlled Text Generation2019 · 509 citations
  4. Transferable Multi-Domain State Generator for Task-Oriented Dialogue Systems2019 · Annual Meeting of the Association for Computational Linguistics (ACL) · 422 citationsWith Pascale Fung and Richard Socher
  5. Mem2Seq: Effectively Incorporating Knowledge Bases into End-to-End Task-Oriented Dialog Systems2018 · Annual Meeting of the Association for Computational Linguistics (ACL) · 307 citationsWith Pascale Fung

From OpenAlex, as of 18 October 2026.

  1. Google alumni founded or led 82 other organizations, the most among the labs we compareIncludes Andrea Madotto. We trace 189 organizations founded or led by AI lab alumni, and 196 career moves between the labs.
  2. The median h-index is 54 at universities and 20 at frontier AI labsIncludes Andrea Madotto. The median h-index is 54 at universities and 20 at frontier AI labs. We compare publishing careers, citations and moves to estimate how universities and frontier labs contribute to AI research.
  3. 14% of AI career moves in the 2020s went from a university to a companyIncludes Andrea Madotto. Our estimates of how university-to-company career transitions changed by decade, including roles added alongside an earlier job.
  4. Google has the highest combined h-index among AI organizationsIncludes Thinking Machines Lab. Our estimates of combined and median h-index at AI organizations, using current affiliations and OpenAlex records.

Colleagues

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

Sources and credits

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

  1. Transferable Multi-Domain State Generator for Task-Oriented Dialogue Systems aclanthology.org. Checked 18 October 2026.
  2. Andrea Madotto andreamad8.github.io. Checked 18 October 2026.
  3. Andrea Madotto | Curriculum andreamad8.github.io. Checked 18 October 2026.
  4. Taming the Beast: Learning to Control Neural Conversational Models arxiv.org. Checked 18 October 2026.
  5. AnyMAL: An Efficient and Scalable Any-Modality Augmented Language Model arxiv.org. Checked 18 October 2026.
  6. Fine-Tuned Language Models Generate Stable Inorganic Materials as Text arxiv.org. Checked 18 October 2026.
  7. PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding proceedings.neurips.cc. Checked 18 October 2026.
  8. Meta's loss is Thinking Machines' gain | TechCrunch techcrunch.com. Checked 18 October 2026.