Philip S. Thomas

111 papers A* 40A 6B 1C 1Journal 53Unranked 10

Year	Rank	Type	Title / Venue / Authors
2025	J	`jnl`	Are Deep Speech Denoising Models Robust to Adversarial Noise? CoRR Will Schwarzer, Philip S. Thomas, Andrea Fanelli, Xiaoyu Liu
2025	J	`jnl`	Fair Representation Learning with Controllable High Confidence Guarantees via Adversarial Inference. CoRR Yuhong Luo, Austin Hoag, Xintong Wang, Philip S. Thomas, Przemyslaw A. Grabowicz
2025	J	`jnl`	Qualia Optimization. CoRR Philip S. Thomas
2025	J	`jnl`	Reinforcement Learning from Human Feedback with High-Confidence Safety Constraints. CoRR Yaswanth Chittepu, Blossom Metevier, Will Schwarzer, Austin Hoag, Scott Niekum, Philip S. Thomas
2025	J	`jnl`	Supervised Reward Inference. CoRR Will Schwarzer, Jordan Schneider, Philip S. Thomas, Scott Niekum
2025	J	`jnl`	Which Rewards Matter? Reward Selection for Reinforcement Learning under Limited Feedback. CoRR Shreyas Chaudhari, Renhao Zhang, Philip S. Thomas, Bruno Castro da Silva
2024	A*	`conf`	Abstract Reward Processes: Leveraging State Abstraction for Consistent Off-Policy Evaluation. NeurIPS Shreyas Chaudhari, Ameet Deshpande, Bruno C. da Silva, Philip S. Thomas
2024	J	`jnl`	Abstract Reward Processes: Leveraging State Abstraction for Consistent Off-Policy Evaluation. CoRR Shreyas Chaudhari, Ameet Deshpande, Bruno Castro da Silva, Philip S. Thomas
2024	—	`conf`	Analyzing the Relationship Between Difference and Ratio-Based Fairness Metrics. FAccT Min-Hsuan Yeh, Blossom Metevier, Austin Hoag, Philip S. Thomas
2024	J	`jnl`	Data-Efficient Policy Evaluation Through Behavior Policy Search. J. Mach. Learn. Res. Josiah P. Hanna, Yash Chandak, Philip S. Thomas, Martha White, Peter Stone, Scott Niekum
2024	A*	`conf`	From Past to Future: Rethinking Eligibility Traces. AAAI Dhawal Gupta, Scott M. Jordan, Shreyas Chaudhari, Bo Liu, Philip S. Thomas, Bruno Castro da Silva
2024	J	`jnl`	ICU-Sepsis: A Benchmark MDP Built from Real Medical Data. RLJ Kartik Choudhary, Dhawal Gupta, Philip S. Thomas
2024	J	`jnl`	ICU-Sepsis: A Benchmark MDP Built from Real Medical Data. CoRR Kartik Choudhary, Dhawal Gupta, Philip S. Thomas
2024	—	`conf`	Low Variance Off-policy Evaluation with State-based Importance Sampling. CAI David M. Bossens, Philip S. Thomas
2024	A*	`conf`	Position: Benchmarking is Limited in Reinforcement Learning Research. ICML Scott M. Jordan, Adam White, Bruno Castro da Silva, Martha White, Philip S. Thomas
2024	J	`jnl`	Position: Benchmarking is Limited in Reinforcement Learning Research. CoRR Scott M. Jordan, Adam White, Bruno Castro da Silva, Martha White, Philip S. Thomas
2024	J	`jnl`	The Cliff of Overcommitment with Policy Gradient Step Sizes. RLJ Scott M. Jordan, Samuel Neumann, James E. Kostas, Adam White, Philip S. Thomas
2023	A	`conf`	Asymptotically Unbiased Off-Policy Policy Evaluation when Reusing Old Data in Nonstationary Environments. AISTATS Vincent Liu, Yash Chandak, Philip S. Thomas, Martha White
2023	J	`jnl`	Asymptotically Unbiased Off-Policy Policy Evaluation when Reusing Old Data in Nonstationary Environments. CoRR Vincent Liu, Yash Chandak, Philip S. Thomas, Martha White
2023	A*	`conf`	Behavior Alignment via Reward Function Optimization. NeurIPS Dhawal Gupta, Yash Chandak, Scott M. Jordan, Philip S. Thomas, Bruno C. da Silva
2023	J	`jnl`	Behavior Alignment via Reward Function Optimization. CoRR Dhawal Gupta, Yash Chandak, Scott M. Jordan, Philip S. Thomas, Bruno Castro da Silva
2023	J	`jnl`	Coagent Networks: Generalized and Scaled. CoRR James E. Kostas, Scott M. Jordan, Yash Chandak, Georgios Theocharous, Dhawal Gupta, Martha White, Bruno Castro da Silva, Philip S. Thomas
2023	J	`jnl`	From Past to Future: Rethinking Eligibility Traces. CoRR Dhawal Gupta, Scott M. Jordan, Shreyas Chaudhari, Bo Liu, Philip S. Thomas, Bruno Castro da Silva
2023	J	`jnl`	Learning Fair Representations with High-Confidence Guarantees. CoRR Yuhong Luo, Austin Hoag, Philip S. Thomas
2023	J	`jnl`	Off-Policy Evaluation for Action-Dependent Non-Stationary Environments. CoRR Yash Chandak, Shiv Shankar, Nathaniel D. Bastian, Bruno Castro da Silva, Emma Brunskill, Philip S. Thomas
2023	J	`jnl`	Optimization using Parallel Gradient Evaluations on Multiple Parameters. CoRR Yash Chandak, Shiv Shankar, Venkata Gandikota, Philip S. Thomas, Arya Mazumdar
2023	—	`conf`	Seldonian Toolkit: Building Software with Safe and Fair Machine Learning. ICSE Companion Austin Hoag, James E. Kostas, Bruno Castro da Silva, Philip S. Thomas, Yuriy Brun
2022	J	`jnl`	Adaptive Rollout Length for Model-Based RL Using Model-Free Deep RL. CoRR Abhinav Bhatia, Philip S. Thomas, Shlomo Zilberstein
2022	J	`jnl`	Enforcing Delayed-Impact Fairness Guarantees. CoRR Aline Weber, Blossom Metevier, Yuriy Brun, Philip S. Thomas, Bruno Castro da Silva
2022	A*	`conf`	Fairness Guarantees under Demographic Shift. ICLR Stephen Giguere, Blossom Metevier, Bruno Castro da Silva, Yuriy Brun, Philip S. Thomas, Scott Niekum
2022	B	`conf`	Mechanizing Soundness of Off-Policy Evaluation. ITP Jared Yeager, J. Eliot B. Moss, Michael Norrish, Philip S. Thomas
2022	A*	`conf`	Off-Policy Evaluation for Action-Dependent Non-stationary Environments. NeurIPS Yash Chandak, Shiv Shankar, Nathaniel D. Bastian, Bruno C. da Silva, Emma Brunskill, Philip S. Thomas
2021	J	`jnl`	Edge-Compatible Reinforcement Learning for Recommendations. CoRR James E. Kostas, Philip S. Thomas, Georgios Theocharous
2021	A*	`conf`	High Confidence Generalization for Reinforcement Learning. ICML James E. Kostas, Yash Chandak, Scott M. Jordan, Georgios Theocharous, Philip S. Thomas
2021	A*	`conf`	High-Confidence Off-Policy (or Counterfactual) Variance Estimation. AAAI Yash Chandak, Shiv Shankar, Philip S. Thomas
2021	J	`jnl`	High-Confidence Off-Policy (or Counterfactual) Variance Estimation. CoRR Yash Chandak, Shiv Shankar, Philip S. Thomas
2021	A	`conf`	Large-scale Interactive Conversational Recommendation System using Actor-Critic Framework. RecSys Ali Montazeralghaem, James Allan, Philip S. Thomas
2021	A*	`conf`	Multi-Objective SPIBB: Seldonian Offline Policy Improvement with Safety Constraints in Finite MDPs. NeurIPS Harsh Satija, Philip S. Thomas, Joelle Pineau, Romain Laroche
2021	J	`jnl`	Multi-Objective SPIBB: Seldonian Offline Policy Improvement with Safety Constraints in Finite MDPs. CoRR Harsh Satija, Philip S. Thomas, Joelle Pineau, Romain Laroche
2021	A*	`conf`	Posterior Value Functions: Hindsight Baselines for Policy Gradient Methods. ICML Chris Nota, Philip S. Thomas, Bruno C. da Silva
2021	A*	`conf`	SOPE: Spectrum of Off-Policy Estimators. NeurIPS Christina J. Yuan, Yash Chandak, Stephen Giguere, Philip S. Thomas, Scott Niekum
2021	J	`jnl`	SOPE: Spectrum of Off-Policy Estimators. CoRR Christina J. Yuan, Yash Chandak, Stephen Giguere, Philip S. Thomas, Scott Niekum
2021	A*	`conf`	Structural Credit Assignment in Neural Networks using Reinforcement Learning. NeurIPS Dhawal Gupta, Gabor Mihucz, Matthew Schlegel, James E. Kostas, Philip S. Thomas, Martha White
2021	A*	`conf`	Towards Practical Mean Bounds for Small Samples. ICML My Phan, Philip S. Thomas, Erik G. Learned-Miller
2021	A*	`conf`	Universal Off-Policy Evaluation. NeurIPS Yash Chandak, Scott Niekum, Bruno C. da Silva, Erik G. Learned-Miller, Emma Brunskill, Philip S. Thomas
2021	J	`jnl`	Universal Off-Policy Evaluation. CoRR Yash Chandak, Scott Niekum, Bruno Castro da Silva, Erik G. Learned-Miller, Emma Brunskill, Philip S. Thomas
2020	A*	`conf`	Asynchronous Coagent Networks. ICML James E. Kostas, Chris Nota, Philip S. Thomas
2020	A*	`conf`	Evaluating the Performance of Reinforcement Learning Algorithms. ICML Scott M. Jordan, Yash Chandak, Daniel Cohen, Mengxue Zhang, Philip S. Thomas
2020	J	`jnl`	Evaluating the Performance of Reinforcement Learning Algorithms. CoRR Scott M. Jordan, Yash Chandak, Daniel Cohen, Mengxue Zhang, Philip S. Thomas
2020	A	`conf`	Is the Policy Gradient a Gradient? AAMAS Chris Nota, Philip S. Thomas
2020	J	`jnl`	Learning Reusable Options for Multi-Task Reinforcement Learning. CoRR Francisco M. Garcia, Chris Nota, Philip S. Thomas
2020	A*	`conf`	Lifelong Learning with a Changing Action Set. AAAI Yash Chandak, Georgios Theocharous, Chris Nota, Philip S. Thomas
2020	A*	`conf`	Optimizing for the Future in Non-Stationary MDPs. ICML Yash Chandak, Georgios Theocharous, Shiv Shankar, Martha White, Sridhar Mahadevan, Philip S. Thomas
2020	J	`jnl`	Optimizing for the Future in Non-Stationary MDPs. CoRR Yash Chandak, Georgios Theocharous, Shiv Shankar, Martha White, Sridhar Mahadevan, Philip S. Thomas
2020	A*	`conf`	Reinforcement Learning When All Actions Are Not Always Available. AAAI Yash Chandak, Georgios Theocharous, Blossom Metevier, Philip S. Thomas
2020	J	`jnl`	Reinforcement Learning for Strategic Recommendations. CoRR Georgios Theocharous, Yash Chandak, Philip S. Thomas, Frits de Nijs
2020	A*	`conf`	Security Analysis of Safe and Seldonian Reinforcement Learning Algorithms. NeurIPS Pinar Ozisik, Philip S. Thomas
2020	A*	`conf`	Towards Safe Policy Improvement for Non-Stationary MDPs. NeurIPS Yash Chandak, Scott M. Jordan, Georgios Theocharous, Martha White, Philip S. Thomas
2020	J	`jnl`	Towards Safe Policy Improvement for Non-Stationary MDPs. CoRR Yash Chandak, Scott M. Jordan, Georgios Theocharous, Martha White, Philip S. Thomas
2019	A	`conf`	A Compression-Inspired Framework for Macro Discovery. AAMAS Francisco M. Garcia, Bruno C. da Silva, Philip S. Thomas
2019	A	`conf`	A Meta-MDP Approach to Exploration for Lifelong Reinforcement Learning. AAMAS Francisco M. Garcia, Philip S. Thomas
2019	A*	`conf`	A Meta-MDP Approach to Exploration for Lifelong Reinforcement Learning. NeurIPS Francisco M. Garcia, Philip S. Thomas
2019	J	`jnl`	A Meta-MDP Approach to Exploration for Lifelong Reinforcement Learning. CoRR Francisco M. Garcia, Philip S. Thomas
2019	J	`jnl`	A New Confidence Interval for the Mean of a Bounded Random Variable. CoRR Erik G. Learned-Miller, Philip S. Thomas
2019	J	`jnl`	Asynchronous Coagent Networks: Stochastic Networks for Reinforcement Learning without Backpropagation or a Clock. CoRR James E. Kostas, Chris Nota, Philip S. Thomas
2019	J	`jnl`	Classical Policy Gradient: Preserving Bellman's Principle of Optimality. CoRR Philip S. Thomas, Scott M. Jordan, Yash Chandak, Chris Nota, James E. Kostas
2019	A*	`conf`	Concentration Inequalities for Conditional Value at Risk. ICML Philip S. Thomas, Erik G. Learned-Miller
2019	J	`jnl`	Is the Policy Gradient a Gradient? CoRR Chris Nota, Philip S. Thomas
2019	A*	`conf`	Learning Action Representations for Reinforcement Learning. ICML Yash Chandak, Georgios Theocharous, James E. Kostas, Scott M. Jordan, Philip S. Thomas
2019	J	`jnl`	Learning Action Representations for Reinforcement Learning. CoRR Yash Chandak, Georgios Theocharous, James E. Kostas, Scott M. Jordan, Philip S. Thomas
2019	J	`jnl`	Lifelong Learning with a Changing Action Set. CoRR Yash Chandak, Georgios Theocharous, Chris Nota, Philip S. Thomas
2019	A*	`conf`	Natural Option Critic. AAAI Saket Tiwari, Philip S. Thomas
2019	A*	`conf`	Offline Contextual Bandits with High Probability Fairness Guarantees. NeurIPS Blossom Metevier, Stephen Giguere, Sarah Brockman, Ari Kobren, Yuriy Brun, Emma Brunskill, Philip S. Thomas
2019	J	`jnl`	Privacy Preserving Off-Policy Evaluation. CoRR Tengyang Xie, Philip S. Thomas, Gerome Miklau
2019	J	`jnl`	Reinforcement Learning When All Actions are Not Always Available. CoRR Yash Chandak, Georgios Theocharous, Blossom Metevier, Philip S. Thomas
2019	J	`jnl`	Reinforcement learning with spiking coagents. CoRR Sneha Aenugu, Abhishek Sharma, Sasikiran Yelamarthi, Hananel Hazan, Philip S. Thomas, Robert Kozma
2018	A*	`conf`	Decoupling Gradient-Like Learning Rules from Representations. ICML Philip S. Thomas, Christoph Dann, Emma Brunskill
2018	A*	`conf`	Importance Sampling for Fair Policy Selection. IJCAI Shayan Doroudi, Philip S. Thomas, Emma Brunskill
2018	J	`jnl`	Natural Option Critic. CoRR Saket Tiwari, Philip S. Thomas
2017	A*	`conf`	Data-Efficient Policy Evaluation Through Behavior Policy Search. ICML Josiah P. Hanna, Philip S. Thomas, Peter Stone, Scott Niekum
2017	J	`jnl`	Data-Efficient Policy Evaluation Through Behavior Policy Search. CoRR Josiah P. Hanna, Philip S. Thomas, Peter Stone, Scott Niekum
2017	J	`jnl`	Decoupling Learning Rules from Representations. CoRR Philip S. Thomas, Christoph Dann, Emma Brunskill
2017	A	`conf`	Importance Sampling for Fair Policy Selection. UAI Shayan Doroudi, Philip S. Thomas, Emma Brunskill
2017	A*	`conf`	Importance Sampling with Unequal Support. AAAI Philip S. Thomas, Emma Brunskill
2017	J	`jnl`	On Ensuring that Intelligent Machines Are Well-Behaved. CoRR Philip S. Thomas, Bruno Castro da Silva, Andrew G. Barto, Emma Brunskill
2017	J	`jnl`	Policy Gradient Methods for Reinforcement Learning with Function Approximation and Action-Dependent Baselines. CoRR Philip S. Thomas, Emma Brunskill
2017	A*	`conf`	Predictive Off-Policy Policy Evaluation for Nonstationary Decision Problems, with Applications to Digital Marketing. AAAI Philip S. Thomas, Georgios Theocharous, Mohammad Ghavamzadeh, Ishan Durugkar, Emma Brunskill
2017	—	`conf`	Using Options and Covariance Testing for Long Horizon Off-Policy Policy Evaluation. NIPS Zhaohan Guo, Philip S. Thomas, Emma Brunskill
2017	J	`jnl`	Using Options for Long-Horizon Off-Policy Evaluation. CoRR Zhaohan Daniel Guo, Philip S. Thomas, Emma Brunskill
2016	A*	`conf`	Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning. ICML Philip S. Thomas, Emma Brunskill
2016	J	`jnl`	Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning. CoRR Philip S. Thomas, Emma Brunskill
2016	A*	`conf`	Energetic Natural Gradient Descent. ICML Philip S. Thomas, Bruno Castro da Silva, Christoph Dann, Emma Brunskill
2016	J	`jnl`	Human-Like Rewards to Train a Reinforcement Learning Controller for Planar Arm Movement. IEEE Trans. Hum. Mach. Syst. Kathleen M. Jagodnik, Philip S. Thomas, Antonie J. van den Bogert, Michael S. Branicky, Robert F. Kirsch
2016	J	`jnl`	Importance Sampling with Unequal Support. CoRR Philip S. Thomas, Emma Brunskill
2016	A*	`conf`	Increasing the Action Gap: New Operators for Reinforcement Learning. AAAI Marc G. Bellemare, Georg Ostrovski, Arthur Guez, Philip S. Thomas, Rémi Munos
2015	J	`jnl`	A Notation for Markov Decision Processes. CoRR Philip S. Thomas
2015	—	`conf`	Ad Recommendation Systems for Life-Time Value Optimization. WWW (Companion Volume) Georgios Theocharous, Philip S. Thomas, Mohammad Ghavamzadeh
2015	A*	`conf`	High Confidence Policy Improvement. ICML Philip S. Thomas, Georgios Theocharous, Mohammad Ghavamzadeh
2015	A*	`conf`	High-Confidence Off-Policy Evaluation. AAAI Philip S. Thomas, Georgios Theocharous, Mohammad Ghavamzadeh
2015	J	`jnl`	Increasing the Action Gap: New Operators for Reinforcement Learning. CoRR Marc G. Bellemare, Georg Ostrovski, Arthur Guez, Philip S. Thomas, Rémi Munos
2015	A*	`conf`	Personalized Ad Recommendation Systems for Life-Time Value Optimization with Guarantees. IJCAI Georgios Theocharous, Philip S. Thomas, Mohammad Ghavamzadeh
2015	—	`conf`	Policy Evaluation Using the Ω-Return. NIPS Philip S. Thomas, Scott Niekum, Georgios Theocharous, George Dimitri Konidaris
2014	A*	`conf`	Natural Temporal Difference Learning. AAAI William Dabney, Philip S. Thomas
2014	J	`jnl`	Proximal Reinforcement Learning: A New Theory of Sequential Decision Making in Primal-Dual Spaces. CoRR Sridhar Mahadevan, Bo Liu, Philip S. Thomas, William Dabney, Stephen Giguere, Nicholas Jacek, Ian Gemp, Ji Liu
2013	—	`conf`	Projected Natural Actor-Critic. NIPS Philip S. Thomas, William Dabney, Stephen Giguere, Sridhar Mahadevan
2012	—	`conf`	Motor primitive discovery. ICDL-EPIROB Philip S. Thomas, Andrew G. Barto
2011	A*	`conf`	Conjugate Markov Decision Processes. ICML Philip S. Thomas, Andrew G. Barto
2011	—	`conf`	Policy Gradient Coagent Networks. NIPS Philip S. Thomas
2011	—	`conf`	TD_gamma: Re-evaluating Complex Backups in Temporal Difference Learning. NIPS George Dimitri Konidaris, Scott Niekum, Philip S. Thomas
2011	A*	`conf`	Value Function Approximation in Reinforcement Learning Using the Fourier Basis. AAAI George Dimitri Konidaris, Sarah Osentoski, Philip S. Thomas
2009	C	`conf`	Application of the Actor-Critic Architecture to Functional Electrical Stimulation Control of a Human Arm. IAAI Philip S. Thomas, Antonie J. van den Bogert, Kathleen M. Jagodnik, Michael S. Branicky