Sheila A. McIlraith
Also known as Sheila McIlraith
Sheila A. McIlraith researches knowledge representation, automated reasoning and sequential decisions. She earned her Toronto PhD in 1997, worked at Stanford and returned to a Toronto faculty role in 2003. Reward machines were introduced at ICML 2018. The shared 2022 JAIR paper develops methods for using reward structure during reinforcement learning, including shaping and task decomposition. It received the shared 2023 IJCAI-JAIR Best Paper Prize.
- Now at
- University of Toronto
- Works
- 240
- Citations
- 14,428
- h-index
- 51
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
1997
Formal accounts of diagnosis
McIlraith received her Toronto computer science PhD in spring 1997, supervised by Raymond Reiter. Her thesis sought a formal account of diagnostic problem solving.
-
1998–2003
Research at Stanford and Toronto
She worked as a researcher at Stanford’s Knowledge Systems Laboratory from January 1998 to September 2003. She became an associate professor at Toronto in November 2003.
-
2018
Reward machines for tasks
Reward machines were introduced at ICML 2018. The work uses automata to represent rewards that depend on the history of a task. It connects a formal task description with the behavior a reinforcement-learning system must learn.
-
2022–2023
Using reward structure in learning
The shared 2022 JAIR paper explains how visible reward structure can support reward shaping, task decomposition and counterfactual reasoning. Its experiments test sample efficiency and policy quality. The paper received the shared IJCAI-JAIR Best Paper Prize in 2023.
Sources and credits
Photo: www.cs.toronto.edu; Sheila A. McIlraith profile photograph; creator not stated. License: Not stated by source. Source: cs.toronto.edu. Cropped to a square and resized. Ask us to remove or replace this photo.