PeopleofSuperintelligence

Researcher

Kriti Aggarwal

Kriti Aggarwal is an AI researcher whose work spans vision-language models, visual document reading, language-model reasoning and healthcare dialogue. She coauthored VLMo in 2021, BEiT-3 in August 2022, DUBLIN in May 2023 and Orca-2 in November 2023. In August 2024, she wrote about healthcare agents that adapt their tone and approach to patients' needs at Hippocratic AI. She coauthored ORION in November 2025 and HEART in January 2026, work on compact reasoning and multi-turn emotional-support dialogue. By March 2026, she is a senior staff scientist at Hippocratic AI.

Timeline

  1. 2021

    Vision-language pretraining with VLMo

    In 2021, she coauthors VLMo, a unified vision-language model whose Mixture-of-Modality-Experts transformer combines modality-specific experts with shared self-attention.

  2. August 2022

    BEiT-3 for vision and language

    In August 2022, she coauthors BEiT-3, a general-purpose multimodal foundation model for vision and vision-language tasks.

  3. May 2023

    DUBLIN visual document model

    In May 2023, she coauthors DUBLIN, a pixel-based model for visual document understanding that is pretrained on web pages with Masked Document Text Generation, Bounding Box and Rendered Question Answering objectives.

  4. November 2023

    Orca-2 small-language reasoning

    In November 2023, she coauthors Orca-2, which teaches small language models to choose different reasoning techniques for different tasks.

  5. August 2024

    Dynamic personalities for care

    In August 2024, she is a Staff Applied Scientist at Hippocratic AI and writes that healthcare agents adapt their tone, style and approach to each patient's context and needs.

  6. November 2025

    Compact reasoning with ORION

    In November 2025, she coauthors ORION, which trains models to reason in compact Mentalese and uses Shorter Length Preference Optimization to reward concise solutions that stay correct.

  7. January 2026

    HEART emotional-support benchmark

    In January 2026, she coauthors HEART, a benchmark that compares human and LLM responses in the same multi-turn emotional-support conversations.

  8. March 2026

    Senior staff scientist

    By March 2026, she is a senior staff scientist at Hippocratic AI.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  2. The United States employs an estimated 48% of people shaping AIWe estimate where the people shaping AI work today, how universities, companies and research institutes split across countries, and how country shares have changed since the 1960s.
  3. 43% of living people shaping AI work outside their country of birthWe estimate that 43% of living people shaping AI work outside their country of birth (520 of 1,198 people with known birth and employment countries).
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionWe estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.

Sources and credits

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

  1. VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts arxiv.org. Checked 21 October 2026.
  2. Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks microsoft.com. Checked 21 October 2026.
  3. DUBLIN -- Document Understanding By Language-Image Network arxiv.org. Checked 21 October 2026.
  4. Orca 2: Teaching Small Language Models How to Reason arxiv.org. Checked 21 October 2026.
  5. The Human Touch in AI: Dynamic Personalities for Better Care hippocraticai.com. Checked 21 October 2026.
  6. ORION: Teaching Language Models to Reason Efficiently in the Language of Thought arxiv.org. Checked 21 October 2026.
  7. HEART: A Unified Benchmark for Assessing Humans and LLMs in Emotional Support Dialogue arxiv.org. Checked 21 October 2026.
  8. AI chatbots now rival human empathy in support conversations earth.com. Checked 21 October 2026.