person

Toby Shevlane

DeepMind model evaluations researcher

DeepMind research scientist focused on dangerous-capability evaluations. Co-authored foundational papers on red-teaming and evaluation frameworks.

current Research Scientist, Model Evaluations, Google DeepMind

@tobyshevlane

Strategy positions

Evals-drivenendorses

Capability/risk evals gate deployment; evals are the load-bearing artefact

Helps design dangerous-capability evaluations and advocates for their adoption as the load-bearing governance artefact.

Dangerous capability evaluations are the minimum viable governance instrument for frontier AI.

§ paperModel evaluation for extreme risks· arXiv· 2023-05· faithful paraphrase

Closest strategy neighbours

by jaccard overlap

Other people whose strategy tags overlap with Toby Shevlane's. Overlap is on tag identity, not stance; opposites can show up if they reference the same tags.

Aleksander Mądry
shared 1 · J=1.00
MIT; ex-OpenAI head of preparedness
Alex Meinke
shared 1 · J=1.00
Apollo Research; deceptive alignment evaluations
Ali Rahimi
shared 1 · J=1.00
Google Brain ML researcher; 'Alchemy' speech
Anna Rogers
shared 1 · J=1.00
IT University of Copenhagen; LLM benchmarking critique
Arati Prabhakar
shared 1 · J=1.00
White House OSTP director (2022–2025)
Beth Barnes
shared 1 · J=1.00
Founder of METR; dangerous capability evaluations

Record last updated 2026-04-25.