PeopleofSuperintelligence

Researcher

Ngân Vũ

Ngân Vũ is an AI researcher whose work spans reinforcement learning, algorithm discovery, compiler optimization and chip design. Her 2023 paper introduced a reinforcement-learning agent for memory mapping. By 2025, she held three Yale degrees and helped develop AlphaEvolve; in 2026, she coauthored Magellan, two ArchAgent papers and a study of mathematicians using AI for discovery. As of October 2026, she is a Science Product Manager at Google DeepMind, where she works on AI for compiler and chip design.

Works
6
Citations
4
h-index
1

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2023

    Reinforcement learning for memory mapping

    In 2023, she coauthors Optimizing Memory Mapping Using Deep Reinforcement Learning, which introduces mallocMuZero, an RL agent that discovers memory mappings that speed up real ML workloads on accelerators.

  2. By 2025

    Three Yale degrees

    By 2025, she holds a B.S./M.S. in computer science and a B.S. in statistics and data science from Yale.

  3. 2025

    AlphaEvolve for algorithm discovery

    In 2025, she helps develop AlphaEvolve, a coding agent that uses an autonomous LLM pipeline to improve algorithms by editing code.

  4. 2026

    Magellan for compiler heuristics

    Magellan, which she coauthors in 2026, combines an LLM coding agent with evolutionary search and autotuning to synthesize executable C++ compiler heuristics.

  5. 2026

    ArchAgent designs cache policies

    ArchAgent, which she coauthors in 2026, builds on AlphaEvolve to design cache-replacement policies. Its policy achieves a 5.3% IPC speedup over prior state of the art on public multi-core Google Workload Traces.

  6. 2026

    Intentmaking and sensemaking

    In 2026, she coauthors Intentmaking and Sensemaking, a paper that reports a study of 11 expert mathematicians who used AlphaEvolve to tackle advanced problems.

  7. 2026

    ArchAgent v2 prefetcher

    ArchAgent v2, which she coauthors in 2026, uses cascaded evolutionary search to design a three-level data prefetcher. Its policy achieves a 3.8% geometric mean IPC speedup over baseline.

  8. As of October 2026

    Science product manager at DeepMind

    As of October 2026, she is a Science Product Manager at Google DeepMind, where she works on AI for compiler and chip design.

Career

  1. Since 2023Google DeepMindScience Product Manager

Education

  1. Yale UniversityB.S. in computer science

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

Most-cited works

  1. Optimizing Memory Mapping Using Deep Reinforcement Learning2023 · 2 citationsWith Julian Schrittwieser and Oriol Vinyals
  2. Magellan: Autonomous Discovery of Novel Compiler Optimization Heuristics with AlphaEvolve2026 · 1 citations
  3. ArchAgent v2: A Case Study with the Data Prefetching Championship2026 · 0 citationsWith Matej Balog
  4. ArchAgent: Agentic AI-driven Computer Architecture Discovery2026 · 0 citationsWith Matej Balog
  5. Intentmaking and Sensemaking: Human Interaction with AI-Guided Mathematical Discovery2026 · 0 citationsWith Adam Zsolt Wagner and Fernanda Viégas

From OpenAlex, as of 21 October 2026.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Ngân Vũ. 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. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  3. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.

Colleagues

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

Sources and credits

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

  1. Optimizing Memory Mapping Using Deep Reinforcement Learning arxiv.org. Checked 21 October 2026.
  2. Speaker: Ngan Vu | K1st World 2022 aisc.events. Checked 21 October 2026.
  3. AlphaEvolve: A Gemini-powered coding agent for designing advanced algorithms deepmind.google. Checked 21 October 2026.
  4. AlphaEvolve: A coding agent for scientific and algorithmic discovery arxiv.org. Checked 21 October 2026.
  5. Magellan: Autonomous Discovery of Novel Compiler Optimization Heuristics with AlphaEvolve arxiv.org. Checked 21 October 2026.
  6. ArchAgent: Agentic AI-driven Computer Architecture Discovery arxiv.org. Checked 21 October 2026.
  7. Intentmaking and Sensemaking: Human Interaction with AI-Guided Mathematical Discovery arxiv.org. Checked 21 October 2026.
  8. ArchAgent v2: A Case Study with the Data Prefetching Championship arxiv.org. Checked 21 October 2026.