Pivotal Spotlight and Farewell Party

​🎓 The 2026 Q3 Pivotal Fellowship is wrapping up, and we're spotlighting research projects for the wider AI safety community starting at 7pm. Come along to hear our fellows present, join the Q&A, and enjoy 🍕 pizza, 🍸 cocktails and snacks.

​🖼️ Posters from the cohort will be up around LISA, so feel free to arrive early.

​✨ Spotlight Talks

​🤖 Luis Montoya — When can language models predict each other? Self- and other-prediction in multi-agent environments
Mentors: Gaël Gendron and Damiano Fornasiere (LawZero)

​🔍 Ethan Nguyen — Localizing Feature Computation in Parameter Space
Mentors: Thomas Dooms (Goodfire) and Logan Riggs (CBAI)

​🧭 Diksha Gupta and Rieke Fruengel — Testing Continuity of Representation Manifolds: Spokes, Not a Ring
Mentor: Stefan Heimersheim (Adecco/GDM)

​🌍 Kacie Yearout — The U.S. Response Calculus: A Framework for Middle-Power Frontier AI Ambitions
Mentors: Robert Trager and Charles Martinet (OMS AIGI)

​🔐 Sleem Abdelghafar — Privacy-Preserving AI Verification via Minimal Information Disclosure
Mentor: Gabriel Kulp (ISL)

​🎭 Ben Maltbie — Persona interactions in long horizon, multi-agent interactions
Mentor: Cozmin Ududec (UK AISI)

​Join Slido for Q&A at https://app.sli.do/event/54CcJAJX8hJHc2jYK7u6Pv