影响力指数

-/100

发表论文2 篇

平均评分3.5

年均产出1.0 篇/年

Elliott Thornley

Postdoc@Massachusetts Institute of Technology·美国·OpenReview

研究方向

AI safety · corrigibility · decision theory · shutdown problem · shutdownability

Towards shutdownable agents: stochastic choice in unseen gridworlds via DReST rewards

ICLR 2026Rejected

Towards shutdownable agents via stochastic choice

ICLR 2025Rejected

合作者 (7)

Alexander Roman

Christos Ziakas

Louis Alexander Thomson