MATS 2026 年秋季项目
2026 年秋季项目为期 10 周，于 9 月 28 日至 12 月 4 日在美国加州伯克利和英国伦敦开展。研究员将接受顶尖研究人员以及 Anthropic、Google DeepMind、OpenAI、Redwood Research 和 ARC 等机构研究人员的指导，并可申请在主项目结束后继续参加 6 至 12 个月的带资助延长期。本届首次开设创业与领域建设方向和生物安全方向。
Applications are now closed.
Program phases
Key dates for the application and admissions timeline
Applications typically open several months before the program begins. Applicants complete a multi-stage admissions process, beginning with a general application. Depending on the tracks, streams, and mentors they apply to, applicants may also complete additional evaluations such as interviews, work tests, coding assessments, or writing samples before final admissions decisions are made.
第一阶段，申请人可申请一个或多个研究方向：实证研究、理论、战略与预测、政策与治理、系统安全、生物安全，以及创业与领域建设。进入第二阶段后，申请人可选择这些方向下的具体研究流，每个研究流由一位或多位拥有各自研究议程的导师负责。你可以查看网格视图。
George Robinson
This stream will focus on monitoring, stress-testing safety methods, and evals, with a focus on risks from scheming AIs. Examples include (black-box) AI control techniques, white-box monitors (probes etc.), chain-of-thought monitoring/faithfulness, building evaluation environments, and stress-testing mitigations.
Advance global DNA synthesis through projects on customer screening, mapping synthesis companies, benchtop integration, and sequence of concern annotation.
This stream focuses on physical access controls for biotech governance: live pathogen repositories, CROs, cloud labs, cell-free expression systems, plasmid vendors, or secondhand lab equipment.
On high-conviction areas, you'll tackle specific open research questions and assess interventions; on exploratory areas, you'll conduct deep dives to determine whether they're worth pursuing. Your findings will directly shape Sentinel's grantmaking strategy and provide strategic guidance to the broader biosecurity community.
Implementing SL4/5 and searching for differentially defense-favored security tools.
We are excited to supervise projects:
- Study the causes and implications of (multi-agent) situational awareness;
- Contribute to LawZero's Scientist AI, in the form of contextualization and uncertainty estimation.
Lee's stream will focus primarily on improving mechanistic interpretability methods for reverse-engineering neural networks.
The SL5 Task Force will build out a prototype SL5 datacenter this year together with frontier AI labs. This will be a massive research and engineering project with many avenues for spinning out new organizations and research programs. This project is urgent due to this technology being needed in the next 1 to 2 years.