PeopleofSuperintelligence

Researcher

Siddharth Mishra-Sharma

Siddharth Mishra-Sharma is a physicist and AI researcher who uses machine learning to study problems in cosmology and astrophysics. He studied Natural Sciences and Mathematics at Peterhouse, Cambridge, earned a Physics PhD at Princeton in 2018, and held research posts at NYU and IAIFI. His work includes neural-likelihood searches for dark matter, self-supervised methods for cosmological data, Maven for supernova science, and audits of language models with hidden objectives. As of October 2026, he is a Member of Technical Staff at Anthropic and an Assistant Professor at Boston University, on leave while working on Claude's scientific capabilities.

Works
96
Citations
2,569
h-index
23
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. By 2018

    Cambridge and Princeton education

    At Peterhouse, University of Cambridge, he reads Natural Sciences (Parts I and II) and Mathematics (Part III), then earns a Physics PhD at Princeton in 2018.

  2. 2018–2021

    NYU and IAIFI research posts

    From 2018 to 2021, he works as a postdoc at NYU's Center for Cosmology and Particle Physics and as an IAIFI Fellow at the NSF AI Institute for Artificial Intelligence and Fundamental Interactions at MIT and Harvard. His research links AI with fundamental physics.

  3. 2021

    Neural likelihood for dark matter

    In 2021, he introduces a neural-likelihood-ratio method using simulation-based inference and neural networks to search astrometric data for gravitational-lensing signatures of dark matter.

  4. 2022

    Machine learning and cosmology

    The 2022 Snowmass white paper "Machine Learning and Cosmology," which he coauthors, surveys machine-learning developments in cosmology and recommends ways to increase their scientific impact.

  5. 2023

    Self-supervised cosmology summaries

    His 2023 paper introduces a self-supervised method with simulation-based augmentations to compress cosmological simulations into summaries for parameter inference.

  6. 2024

    Maven for supernova science

    Maven, which he coauthors in 2024, is a foundation model for supernova science that aligns photometric and spectroscopic observations.

  7. 2025

    Auditing hidden objectives

    In 2025, he coauthors an alignment-audit study that trains a language model with a hidden objective and tests whether audit teams can detect it.

  8. As of October 2026

    Anthropic and BU appointments

    As of October 2026, he is a Member of Technical Staff at Anthropic, where he works on Claude's scientific capabilities, and an Assistant Professor of Computing & Data Sciences and Physics at Boston University, currently on leave. In March 2026, he explains how to apply multi-day agentic coding workflows to scientific computing as a researcher on Anthropic's Discovery team.

Career

  1. Since 2026Boston UniversityAssistant Professor of Computing & Data Sciences and Physics
  2. Since 2024AnthropicMember of Technical Staff
  3. 2022–2026Harvard University
  4. 2022–2026Massachusetts Institute of Technology
  5. 2018–2021New York UniversityPostdoc
  6. 2018–2021Institute for Artificial Intelligence and Fundamental InteractionsIAIFI Fellow

Education

  1. 2018Princeton UniversityPhD
  2. University of Cambridge

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

Most-cited works

  1. Low-Mass Dark Matter Search with the DarkSide-50 Experiment2018 · Physical Review Letters · 442 citations
  2. Searching for long-lived particles beyond the Standard Model at the Large Hadron Collider2020 · Journal of Physics G Nuclear and Particle Physics · 263 citations
  3. Constraints on Sub-GeV Dark-Matter–Electron Scattering from the DarkSide-50 Experiment2018 · Physical Review Letters · 253 citations
  4. Modulation effects in dark matter-electron scattering experiments2015 · Physical review. D. Particles, fields, gravitation, and cosmology/Physical review. D. Particles and fields · 149 citations
  5. LHC searches for dark sector showers2017 · Journal of High Energy Physics · 120 citations

From OpenAlex, as of 21 October 2026.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Siddharth Mishra-Sharma. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  2. 16% of AI career moves in the 2020s went from a university to a companyIncludes Siddharth Mishra-Sharma. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  3. 22 of 133 people shaping AI with a former OpenAI role now work at AnthropicIncludes Anthropic. Our estimates of where OpenAI recruits came from and where 133 people with former OpenAI roles work now, including the companies and labs they founded.
  4. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes Anthropic. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.

Colleagues

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

Sources and credits

Photo: We have no published photo of Siddharth Mishra-Sharma yet. Send us a free-licensed photo.

  1. Siddharth Mishra-Sharma | Cosmology sites.bu.edu. Checked 21 October 2026.
  2. [2110.01620] Inferring dark matter substructure with astrometric lensing beyond the power spectrum arxiv.org. Checked 21 October 2026.
  3. [2203.08056] Machine Learning and Cosmology arxiv.org. Checked 21 October 2026.
  4. [2308.09751] Data Compression and Inference in Cosmology with Self-Supervised Machine Learning arxiv.org. Checked 21 October 2026.
  5. [2408.16829] Maven: A Multimodal Foundation Model for Supernova Science arxiv.org. Checked 21 October 2026.
  6. [2503.10965] Auditing language models for hidden objectives arxiv.org. Checked 21 October 2026.
  7. Siddharth Mishra-Sharma smsharma.io. Checked 21 October 2026.
  8. Long-running Claude for scientific computing anthropic.com. Checked 21 October 2026.