Christoph Dann

Theoretical Foundations and Effective Algorithms for Policy-Aware Simulator Learning
Christoph Dann, Yishay Mansour, Mehryar Mohri
[Arxiv]

Principled Model Routing for Unknown Mixtures of Source Domains
Christoph Dann, Yishay Mansour, Teodor V. Marinov, Mehryar Mohri
NeurIPS '25 [OpenReview] [Poster]

Mitigating Preference Hacking in Policy Optimization with Pessimism
Dhawal Gupta, Adam Fisch, Christoph Dann, Alekh Agarwal
[Arxiv]

Can RLHF be More Efficient with Imperfect Reward Models? A Policy Coverage Perspective
Jiawei Huang, Bingcong Li, Christoph Dann, Niao He
ICML '25 [pdf] [Arxiv]

Design Considerations in Offline Preference-based RL
Alekh Agarwal, Christoph Dann, Teodor V. Marinov
ICML '25 [pdf] [Arxiv]

Rate-Preserving Reductions for Blackwell Approachability
Christoph Dann, Yishay Mansour, Mehryar Mohri, Jon Schneider, Balasubramanian Sivan
COLT '25 [pdf] [Arxiv]

Preserving Expert-Level Privacy in Offline Reinforcement Learning
Navodita Sharma, Vishnu Vinod, Abhradeep Thakurta, Alekh Agarwal, Borja Balle, Christoph Dann, Aravindan Raghuveer
TMLR '25 J2C certification [Arxiv]

Also presented in the Journal-to-Conference Track at ICML '26

Conditional Language Policy: A General Framework for Steerable Multi-Objective Finetuning
Kaiwen Wang, Rahul Kidambi, Ryan Sullivan, Alekh Agarwal, Christoph Dann, Andrea Michi, Marco Gelmi, Yunxuan Li, Raghav Gupta, Avinava Dubey, Alexandre Ramé, Johan Ferret, Geoffrey Cideron, Le Hou, Hongkun Yu, Amr Ahmed, Aranyak Mehta, Léonard Hussenot, Olivier Bachem, Edouard Leurent
Findings of EMNLP '24 [pdf] [Arxiv]

A Minimaximalist Approach to Reinforcement Learning from Human Feedback
Gokul Swamy, Christoph Dann, Rahul Kidambi, Zhiwei Steven Wu, Alekh Agarwal
ICML '24 [pdf] [Arxiv]

Data-Driven Online Model Selection With Regret Guarantees
Aldo Pacchiano, Christoph Dann, Claudio Gentile
AISTATS '24 [Arxiv]

A Blackbox Approach to Best of Both Worlds in Bandits and Beyond
Christoph Dann, Chen-Yu Wei, Julian Zimmert
COLT '23 [Arxiv]

Learning in POMDPs is Sample-Efficient with Hindsight Observability
Jonathan N. Lee, Alekh Agarwal, Christoph Dann, Tong Zhang
ICML '23 [Arxiv]

Reinforcement Learning Can Be More Efficient with Multiple Rewards
Christoph Dann, Yishay Mansour, Mehryar Mohri
ICML '23 [pdf]

Best of Both Worlds Policy Optimization
Christoph Dann, Chen-Yu Wei, Julian Zimmert
ICML '23 short live talk [pdf] [Arxiv]

Multiple-policy High-confidence Policy Evaluation
Christoph Dann, Mohammad Ghavamzadeh, Teodor V. Marinov
AISTATS '23 [pdf]

Pseudonorm Approachability and Applications to Regret Minimization
Christoph Dann, Yishay Mansour, Mehryar Mohri, Jon Schneider, Balasubramanian Sivan
ALT '23 [Arxiv]

A Unified Algorithm for Stochastic Path Problems
Christoph Dann, Chen-Yu Wei, Julian Zimmert
ALT '23 [Arxiv]

Best of Both Worlds Model Selection
Aldo Pacchiano, Christoph Dann, Claudio Gentile
NeurIPS '22 [Arxiv]

Guarantees for Epsilon-Greedy Reinforcement Learning with Function Approximation
Christoph Dann, Yishay Mansour, Mehryar Mohri, Ayush Sekhari, Karthik Sridharan
ICML '22 [Arxiv]

A Model Selection Approach for Corruption Robust Reinforcement Learning
Chen-Yu Wei, Christoph Dann, Julian Zimmert
ALT '22 best paper award [Arxiv]

Leveraging Initial Hints for Free in Stochastic Linear Bandits
Richard Zhang, Abhimanyu Das, Ashok Cutkosky, Christoph Dann
ALT '22 [Arxiv]

Same Cause; Different Effect in the Brain
Mariya Toneva, Jennifer Williams, Anand Bollu, Christoph Dann, Leila Wehbe
CLeaR '22 [Arxiv]

Agnostic Reinforcement Learning with Low-Rank MDPs and Rich Observations
Christoph Dann, Yishay Mansour, Mehryar Mohri, Ayush Sekhari, Karthik Sridharan
NeurIPS '21 with spotlight talk [Arxiv]

Beyond Value-Function Gaps: Improved Instance-Dependent Regret Bounds for Episodic Reinforcement Learning
Christoph Dann, Teodor V. Marinov, Mehryar Mohri, Julian Zimmert
NeurIPS '21 with spotlight talk [Arxiv]

A Provably Efficient Model-Free Posterior Sampling Method for Episodic Reinforcement Learning
Christoph Dann, Mehryar Mohri, Tong Zhang, Julian Zimmert
NeurIPS '21 [Arxiv]

Neural Active Learning with Performance Guarantees
Pranjal Awasthi, Christoph Dann, Claudio Gentile, Ayush Sekhari, Zhilei Wang
NeurIPS '21 [Arxiv]

Dynamic Balancing for Model Selection in Bandits and RL
Ashok Cutkosky, Christoph Dann, Abhimanyu Das, Claudio Gentile, Aldo Pacchiano, Manish Purohit
ICML '21 [pdf]

This paper is based on this earlier report:
Regret Bound Balancing and Elimination for Model Selection in Bandits and RL
Aldo Pacchiano, Christoph Dann, Claudio Gentile, Peter Bartlett, 2020
[Arxiv]

Reinforcement Learning with Feedback Graphs
Christoph Dann, Yishay Mansour, Mehryar Mohri, Ayush Sekhari, Karthik Sridharan
NeurIPS '20 [pdf] [Arxiv]

Scaling up behavioral science interventions in online education
René F. Kizilcec, Justin Reich, Michael Yeomans, Christoph Dann, Emma Brunskill, Glenn Lopez, Selen Turkay, Joseph Jay Williams, Dustin Tingley
PNAS '20 [pdf]

Being Optimistic to Be Conservative: Quickly Learning a CVaR Policy
Ramtin Keramati, Christoph Dann, Alex Tamkin, Emma Brunskill
AAAI '20 [Arxiv]

Policy Certificates: Towards Accountable Reinforcement Learning
Christoph Dann, Lihong Li, Wei Wei, Emma Brunskill
ICML '19 [Arxiv]

Also presented at the Workshop on Ethical, Social and Governance Issues in AI at NeurIPS '18

On Oracle-Efficient PAC Reinforcement Learning with Rich Observations
Christoph Dann, Nan Jiang, Akshay Krishnamurthy, Alekh Agarwal, John Langford, Robert Schapire
NeurIPS '18 with spotlight talk [Arxiv]

Previously presented at 12th Annual Machine Learning Symposium at NYAS

Decoupling Gradient-Like Learning Rules from Representations
Philip S. Thomas, Christoph Dann, Emma Brunskill
ICML '18 [Arxiv]

Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learning
Christoph Dann, Tor Lattimore, Emma Brunskill
NeurIPS '17 with spotlight talk [pdf] [Arxiv] [code]

Previous version of this paper with stronger focus on linear state space dependency in PAC bound:
UBEV - A More Practical Algorithm for Episodic RL with Near-Optimal PAC and Regret Guarantees
Christoph Dann, Tor Lattimore, Emma Brunskill, March 2017
[Arxiv]

Sample Efficient Policy Search for Optimal Stopping Domains
Karan Goel, Christoph Dann, Emma Brunskill
IJCAI '17 [Arxiv]

Automated Matching of Pipeline Corrosion Features from In-line Inspection Data
Markus Dann, Christoph Dann
Reliability Engineering and System Safety '17 [pdf]

Memory Lens - How Much Memory Does an Agent Use?
Christoph Dann, Katja Hofmann, Sebastian Nowozin
EWRL '16 [pdf] [Arxiv]

Also presented at the Interpretable ML for Complex Systems Workshop at NeurIPS '16

Energetic Natural Gradient Descent
Philip S. Thomas, Bruno Castro da Silva, Christoph Dann, Emma Brunskill
ICML '16 [pdf]

Bayesian Time-of-Flight for Realtime Shape, Illumination and Albedo
Amit Adam, Christoph Dann, Omer Yair, Shai Mazor, Sebastian Nowozin
PAMI '16 [pdf] [pdf supp] [video] [publisher pdf] [Arxiv]

Sample Complexity of Episodic Fixed-Horizon Reinforcement Learning
Christoph Dann, Emma Brunskill
NeurIPS '15 [Arxiv]

The Human Kernel
Andrew Gordon Wilson, Christoph Dann, Christopher G. Lucas, Eric P. Xing
NeurIPS '15 [pdf] [Arxiv]

Thoughts on Massively Scalable Gaussian Processes
Andrew Gordon Wilson, Christoph Dann, Hannes Nickisch, 2015
[Arxiv]

RLPy: A Value-Function-Based Reinforcement Learning Framework for Education and Research
Alborz Geramifard, Christoph Dann, Robert H. Klein, William Dabney, Jonathan P. How
JMLR MLOSS Track '15 [pdf] [code]

Policy Evaluation with Temporal Differences: A Survey and Comparison
Christoph Dann, Gerhard Neumann, Jan Peters
JMLR '14 [pdf] [N-link pendulum linearization derivation] [code]

Also presented at ICAPS 2015 [pdf]

Off-Policy Learning Combined with Automatic Feature Expansion for Solving Large MDPs
Alborz Geramifard, Christoph Dann, Jonathan P. How
RLDM '13 [pdf] [poster]

Pottics - The Potts Topic Model for Semantic Image Segmentation
Christoph Dann, Peter Gehler, Stefan Roth, Sebastian Nowozin
DAGM '12 [pdf] [publisher pdf] [poster] [code]

Christoph Dann

Publications and Preprints

Misc