PeopleofSuperintelligence

Researcher

Adam Yang

Also known as Adam X. Yang, adamxyang

Adam Yang is a machine-learning researcher whose work spans kernel methods, Bayesian language models and AI reasoning. He coauthored papers on deep kernel processes and Bayesian representation learning and later coauthored work on language-model calibration and reward-model alignment. Magistral is Mistral's first reasoning model and Devstral-Small is an open model for code agents; he coauthored both papers in 2025. Reasoning Arena routes non-diverse reward groups to a judge system; he coauthored the paper in June 2026. As of October 2026, he is an AI Scientist at Mistral AI.

Works
7
Citations
20
h-index
3
Badges

Publication numbers from OpenAlex, as of 18 October 2026.

Timeline

  1. 2021

    Deep kernel processes

    In 2021, he coauthors Deep kernel processes, which progressively transforms positive definite Gram matrices with nonlinear kernels and inverse Wishart sampling.

  2. 2023

    Bayesian representation learning

    “A theory of representation learning gives a deep generalisation of kernel methods” introduces the Bayesian representation learning limit; Yang coauthors the 2023 paper.

  3. 2024

    Calibrating language models

    In 2024, he coauthors Laplace-LoRA, a Bayesian method that applies a Laplace approximation to LoRA parameters and improves calibration in fine-tuned language models. Bayesian Reward Models for LLM Alignment is a 2024 paper he coauthors; its uncertainty estimates mitigate reward overoptimization in best-of-n sampling.

  4. June 2025

    Mistral's first reasoning model

    Magistral is Mistral's first reasoning model; Yang coauthors the June 2025 report on the model and its scalable reinforcement-learning pipeline.

  5. 2025

    Models for code agents

    Devstral-Small is a lightweight open-source model for code agents introduced in the Devstral paper in 2025; Yang is a coauthor.

  6. June 2026

    Reasoning Arena

    Reasoning Arena is an adaptive training framework that routes non-diverse reward groups to a judge system; Yang coauthors the paper in June 2026.

  7. As of October 2026

    AI Scientist at Mistral AI

    As of October 2026, he is an AI Scientist at Mistral AI.

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

Career

  1. Since 2026Mistral AIAI Scientist
  2. 2020–2024University of Bristol

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

Most-cited works

  1. Bayesian Low-rank Adaptation for Large Language Models2023 · 6 citations
  2. Deep kernel processes2020 · 5 citations
  3. Instruction Tuning With Loss Over Instructions2024 · neural information processing systems · 4 citationsWith Emine Yilmaz
  4. Bayesian Reward Models for LLM Alignment2024 · 2 citations
  5. MONGOOSE: Path-wise Smooth Bayesian Optimisation via Meta-learning2023 · 2 citations

From OpenAlex, as of 18 October 2026.

  1. The median h-index is 55 at universities and 20 at frontier AI labsIncludes Adam Yang. 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.
  2. Google alumni founded or led 77 other organizations, the most among the labs we compareIncludes Mistral AI. We trace 177 organizations founded or led by AI lab alumni, and 186 career moves between the labs.
  3. The company share of AI researchers’ top papers rose from 2% to 33%Includes Mistral AI. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 2% in 1975–1979 to 33% in 2020–2024.
  4. Google has the highest combined h-index among AI organizationsIncludes Mistral AI. 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 Adam Yang yet. Send us a free-licensed photo.

  1. Deep kernel processes proceedings.mlr.press. Checked 18 October 2026.
  2. A theory of representation learning gives a deep generalisation of kernel methods proceedings.mlr.press. Checked 18 October 2026.
  3. Bayesian Low-rank Adaptation for Large Language Models proceedings.iclr.cc. Checked 18 October 2026.
  4. Bayesian Reward Models for LLM Alignment arxiv.org. Checked 18 October 2026.
  5. Magistral arxiv.org. Checked 18 October 2026.
  6. Devstral: Fine-tuning Language Models for Coding Agent Applications arxiv.org. Checked 18 October 2026.
  7. adamxyang (Adam Yang) · GitHub github.com. Checked 18 October 2026.
  8. Reasoning Arena: Trace Tournaments When Verifiable Rewards Fall Short arxiv.org. Checked 18 October 2026.