MATS Summer 2022
The Summer 2022 cohort was MATS's first full-scale program, with 31 fellows and 7 mentors from leading AI safety organizations including OpenAI, ARC, MIRI, EleutherAI, and Aligned AI. The program ran 5 weeks online followed by 8 weeks in-person in Berkeley, where fellows conducted independent research under expert mentorship, participated in educational seminars, and built community with peers in the Berkeley AI safety ecosystem.
Applications are now open. Apply by June 7th.
Program phases
Key dates for the application and admissions timeline
General Application (May 12th to June 7th)
Applicants fill out a general application to individual tracks which should take 1-2 hours. Applications are due by June 7th EOD AOE.
Additional Evaluations (June 7th to late July)
After an initial evaluation, applicants will apply to individual streams listed below. Additionally, applicants undergo a variety of track specific evaluations including coding tests, writing reviews, work tests, and interviews. Which evaluations you will undergo depend on the tracks, streams and mentors you apply to.
Admissions Decisions (Late July to early August)
Selected applicants are notified of their acceptance and anticipated mentor later in the application cycle.
Related research
SolidGoldMagikarp (plus, prompt generation)
Anomalous tokens: a mysterious failure mode for GPT (which reliably insulted Matthew)
- We have found a set of anomalous tokens which result in a previously undocumented failure mode for GPT-2 and GPT-3 models. (The 'instruct' models “are particularly deranged” in this context, as janus has observed.)
- Many of these tokens reliably break determinism in the OpenAI GPT-3 playground at temperature 0 (which theoretically shouldn't happen).
Authors:
Jessica Cooper (Rumbelow), Matthew Watkins
Jessica Cooper, Matthew Watkins
Date:
Apr 23, 2026
Citations:
18
A Toy Model of Universality: Reverse Engineering How Networks Learn Group Operations
Universality is a key hypothesis in mechanistic interpretability -- that different models learn similar features and circuits when trained on similar tasks. In this work, we study the universality hypothesis by examining how small neural networks learn to implement group composition. We present a novel algorithm by which neural networks may implement composition for any finite group via mathematical representation theory. We then show that networks consistently learn this algorithm by reverse engineering model logits and weights, and confirm our understanding using ablations. By studying networks of differing architectures trained on various groups, we find mixed evidence for universality: using our algorithm, we can completely characterize the family of circuits and features that networks learn on this task, but for a given network the precise circuits learned -- as well as the order they develop -- are arbitrary.
Authors:
Bilal Chughtai
Bilal Chughtai, Lawrence Chan, Neel Nanda
Date:
May 7, 2026
Citations:
144
The MATS Program is supported by a diverse and highly respected group of mentors — top-tier researchers, engineers, and thinkers working across AI alignment, governance, interpretability, and security.
Community at MATS
MATS Research phase provides scholars with a community of peers.
Scholars work out of a shared office and are supported by the Community Team.
MATS alumni report that the connections with peers that they made during MATS have had the largest impact on them years later. Our full-time Community Team works to facilitate these connections and also provide general well-being support. Weekly lightning talks, scholar-led discussion groups, game nights, and outings to SF are some examples of MATS events.