PeopleofSuperintelligence

Researcher

Abhinav Rastogi

As of February 2026, Abhinav Rastogi is a researcher at Mistral AI. His research included scalable dialogue state tracking, virtual-assistant datasets and programmable dialogue systems. In 2024, he co-authored a study that found RLAIF achieved comparable performance to RLHF across summarization, helpful dialogue and harmless dialogue. In 2025, he co-authored Magistral, Mistral's first reasoning model, and in 2026 he co-authored the Ministral 3 model family.

Works
49
Citations
2,504
h-index
20
Badges

Publication numbers from OpenAlex, as of 18 October 2026.

Timeline

  1. 2017

    Scaling dialogue state tracking

    In 2017, he co-authors a framework for multi-domain dialogue state tracking that represents each slot with a bounded candidate set.

  2. 2019

    A schema-guided dialogue dataset

    In 2019, he and Pranav Khaitan introduce the Schema-Guided Dialogue dataset for virtual assistants. Their approach uses natural-language service descriptions so one model can work with services absent from training.

  3. 2023

    Programming task-oriented dialogue

    He co-authors AnyTOD, a programmable system for task-oriented dialogue. In 2023, the paper reports strong results on the STAR, ABCD and SGD benchmarks, plus zero-shot transfer to MultiWOZ.

  4. 2024

    AI feedback for alignment

    In 2024, he co-authors RLAIF vs. RLHF, a study of reinforcement learning from AI feedback. Its results are comparable to RLHF across summarization, helpful dialogue and harmless dialogue.

  5. June 2025

    Mistral's first reasoning model

    In June 2025, he co-authors Magistral, which Mistral introduces as its first reasoning model and scalable reinforcement-learning pipeline.

  6. January 2026

    The Ministral 3 model family

    In January 2026, he co-authors Ministral 3, a family of parameter-efficient language models in three sizes, with image understanding and reasoning variants.

  7. February 2026

    Research at Mistral AI

    In February 2026, he speaks as part of Mistral AI's research team. His education includes IIT Bombay and an MS from Stanford.

A language model matched the links in this timeline to people, organizations and papers (provisional, model-judged).

Career

  1. NowMistral AIResearcher
  2. 2022–2025Google DeepMind
  3. 2017–2023Google

Education

  1. IIT Bombay
  2. Stanford UniversityMS

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

Most-cited works

  1. Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models2022 · 569 citationsWith Abubakar Abid, Chris Callison-Burch, Christopher D. Manning, Christopher Potts, Dan Roth and Danqi Chen
  2. Towards Scalable Multi-Domain Conversational Agents: The Schema-Guided Dialogue Dataset2020 · 456 citations
  3. A Fast Unified Model for Parsing and Sentence Understanding2016 · Annual Meeting of the Association for Computational Linguistics (ACL) · 229 citationsWith Christopher D. Manning, Christopher Potts and Samuel R. Bowman
  4. MultiWOZ 2.2: A Dialogue Dataset with Additional Annotation Corrections and State Tracking Baselines2020 · 188 citations
  5. Building a Conversational Agent Overnight with Dialogue Self-Play2018 · 147 citations

From OpenAlex, as of 18 October 2026.

  1. Google alumni founded or led 77 other organizations, the most among the labs we compareIncludes Abhinav Rastogi. We trace 177 organizations founded or led by AI lab alumni, and 186 career moves between the labs.
  2. The median h-index is 55 at universities and 20 at frontier AI labsIncludes Abhinav Rastogi. The median h-index is 55 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. Of 401 people shaping AI who worked at Google, 157 work there nowIncludes Abhinav Rastogi. Our estimates of where former Google and Google DeepMind people work now, including moves within Google.
  4. Google has the highest combined h-index among AI organizationsIncludes Google. 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 Abhinav Rastogi yet. Send us a free-licensed photo.

  1. Scalable Multi-Domain Dialogue State Tracking research.google. Checked 18 October 2026.
  2. Introducing the Schema-Guided Dialogue Dataset for Conversational Assistants research.google. Checked 18 October 2026.
  3. AnyTOD: A Programmable Task-Oriented Dialog System aclanthology.org. Checked 18 October 2026.
  4. RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback proceedings.mlr.press. Checked 18 October 2026.
  5. Magistral arxiv.org. Checked 18 October 2026.
  6. Ministral 3 arxiv.org. Checked 18 October 2026.
  7. Innovating the Future with AI gymkhana.iitb.ac.in. Checked 18 October 2026.