Join our team
Work with us directly to advance AI safety research. No roles are open at the moment. If you think you may be a good fit for the team, introduce yourself and we will reach out when something opens up.
Research
A selection of research conducted by Pivotal fellows.
1 Aug 2026
A False Average: Chain-of-Thought Monitors Collapse Where They Are the Only Defense
Researcher: Shikhar Shiromani
Mentor: Noah Y. Siegel
Technical AI Safety
6 Jul 2026
Compressed Computation under L⁴ Loss is likely Computation in Superposition
Researcher: Francisco Ferreira da Silva
Mentor: Stefan Heimersheim
Technical AI Safety
3 Jul 2026
Individual Parameters in Weight-Sparse Transformers Appear Interpretable
Researcher: Arnau Marin-Llobet
Mentor: Stefan Heimersheim
Technical AI Safety
24 Jun 2026
Door's Locked, Try the Window: Measuring Circumvention Propensity in Coding Agents
Researcher: Prakrat Agrawal
Mentor: Jérémy Scheurer
Technical AI Safety
23 Jun 2026
Evidence for Feature-Specific Error Correction in LLMs
Researcher: Francisco Ferreira da Silva
Mentor: Stefan Heimersheim
Technical AI Safety
14 May 2026
Training on Documents About Monitoring Leads to CoT Obfuscation
Researcher: Reilly Haskins
Mentors: Joshua Engels & Bilal Chughtai
Technical AI Safety
30 Apr 2026
Causal Foundations of Collective Agency
Researcher: Frederik Hytting Jørgensen
Mentor: Lewis Hammond
Technical AI Safety
23 Apr 2026
Bayesian Influence Functions for Hessian-Free Data Attribution
Researcher: Philipp Alexander Kreer
Mentor: Jesse Hoogland
Technical AI Safety
7 Apr 2026
Stochastic Gradient Descent in the Saddle-to-Saddle Regime of Deep Linear Networks
Researchers: Guillaume Corlouer & Avi Semler
Mentors: Alexander Strang & Alexander Gietelink Oldenziel
Technical AI Safety
4 Feb 2026
Attack Selection Reduces Safety in Concentrated AI Control Settings against Trusted Monitoring
Researchers: Joachim Schaeffer & Arjun Khandelwal
Mentor: Tyler Tracy
Technical AI Safety
1 Dec 2025
Factor(T,U): Factored Cognition Strengthens Monitoring of Untrusted AI
Researcher: Aaron Sandoval
Mentor: Cody Rushing
Technical AI Safety
1 Dec 2025
Beyond Vibe Decision Theory: Asymmetric Manipulation Vulnerabilities in LLM Multi-Agent Coordination
Researcher: Sukanya Krishna
Mentor: Tobin South
Technical AI Safety
20 Nov 2025
The Multi-Agent Off-Switch Game
Researcher: Soroush Ebadian
Mentor: Lewis Hammond
Technical AI Safety
12 Nov 2025
Decomposition of Small Transformer Models
Researcher: Casper L. Christensen
Mentor: Logan Riggs Smith
Technical AI Safety
14 Oct 2025
Influence Dynamics and Stagewise Data Attribution
Researchers: Jin Hwa Lee & Matt Smith
Mentor: Jesse Hoogland
Technical AI Safety
30 Sep 2025
Activation Probes Are Reliable With a Handful of Positive Examples
Researcher: Riya Tyagi
Mentor: Stefan Heimersheim
Technical AI Safety
26 Jul 2025
Trivial Trojans: How Minimal MCP Servers Enable Cross-Tool Exfiltration of Sensitive Data
Researcher: Nicola Croce
Mentor: Tobin South
Technical AI Safety
16 Jul 2025
Benchmarking Deception Probes via Black-to-White Performance Boosts
Researcher: Avi Parrack
Mentor: Stefan Heimersheim
Technical AI Safety
26 May 2025
Understanding the learned look-ahead behavior of chess neural networks
Researcher: Diogo Cruz
Technical AI Safety
Get involved
Mentor top emerging researchers
Guide a fellow on a research project, share your expertise, and help them grow as a researcher. We handle recruitment, logistics, and day-to-day research support.
Participate in a fellowship
Join our in-person research fellowship in London and work with experienced mentors on AI safety, AI governance, AIxBio, or biodefense.