Özgür Şimşek — accepted AI-conference papers

2026

CausalProfiler: Generating Synthetic Benchmarks for Rigorous and Transparent Evaluation of Causal Machine Learning

Panayiotis Panayiotou, Audrey Poinsot, Alessandro Leite, Nicolas CHESNEAU, Marc Schoenauer, Özgür Şimşek

ICML 2026poster

Causal machine learning (Causal ML) aims to answer "what if" questions using machine learning algorithms, making it a promising tool for high-stakes decision-making. Yet, empirical evaluation practices in Causal ML remain limited. Existing benchmarks often rely on a handful of hand-crafted or semi-s…

Cited by 0 Source ↗Scholar ↗

2026

Learning the Minimum Action Distance

Lorenzo Steccanella, Joshua B. Evans, Özgür Şimşek, Anders Jonsson

ICML 2026poster

This paper presents a state representation framework for Markov decision processes (MDPs) that can be learned solely from state trajectories, requiring neither reward signals nor the actions executed by the agent. We propose learning the $\textit{minimum action distance}$ (MAD), defined as the minim…

Cited by 0 Source ↗Scholar ↗

2025

Accelerating Task Generalisation with Multi-Level Skill Hierarchies

Thomas P Cannon, Özgür Şimşek

ICLR 2025poster

Developing reinforcement learning agents that can generalise effectively to new tasks is one of the main challenges in AI research. This paper introduces Fracture Cluster Options (FraCOs), a multi-level hierarchical reinforcement learning method designed to improve generalisation performance. FraCOs…

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

Approximating Shapley Explanations in Reinforcement Learning

Daniel Beechey, Özgür Şimşek

NeurIPS 2025poster

Reinforcement learning has achieved remarkable success in complex decision-making environments, yet its lack of transparency limits its deployment in practice, especially in safety-critical settings. Shapley values from cooperative game theory provide a principled framework for explaining reinforcem…

Cited by 0 Source ↗Scholar ↗

2025

Novel Exploration via Orthogonality

Andreas Theophilou, Özgür Şimşek

NeurIPS 2025poster

Efficient exploration remains one of the key open problems in reinforcement learning. Discovering novel states or transitions efficiently requires policies that effectively direct the agent away from regions of the state space that are already well explored. We introduce Novel Exploration via Orthog…

Cited by 0 Source ↗Scholar ↗

2025

Position: Causal Machine Learning Requires Rigorous Synthetic Experiments for Broader Adoption

Audrey Poinsot, Panayiotis Panayiotou, Alessandro Leite, Nicolas CHESNEAU, Özgür Şimşek, Marc Schoenauer

ICML 2025poster

Causal machine learning has the potential to revolutionize decision-making by combining the predictive power of machine learning algorithms with the theory of causal inference. However, these methods remain underutilized by the broader machine learning community, in part because current empirical ev…

Cited by 0 Source ↗PDF ↗Scholar ↗

2023

Creating Multi-Level Skill Hierarchies in Reinforcement Learning

Joshua Benjamin Evans, Özgür Şimşek

NeurIPS 2023poster

What is a useful skill hierarchy for an autonomous agent? We propose an answer based on a graphical representation of how the interaction between an agent and its environment may unfold. Our approach uses modularity maximisation as a central organising principle to expose the structure of the intera…

Cited by 4 Source ↗PDF ↗code ↗Scholar ↗

2023

Explaining Reinforcement Learning with Shapley Values

Daniel Beechey, Thomas M. S. Smith, Özgür Şimşek

ICML 2023poster

For reinforcement learning systems to be widely adopted, their users must understand and trust them. We present a theoretical analysis of explaining reinforcement learning using Shapley values, following a principled approach from game theory for identifying the contribution of individual players to…

Cited by 24 Source ↗PDF ↗code ↗Scholar ↗

2023

Resource-Constrained Station-Keeping for Latex Balloons Using Reinforcement Learning

Jack Saunders, Loïc Prenevost, Özgür Şimşek, Alan Hunter, Wenbin Li

IROS 2023poster

High altitude balloons have proved useful for ecological aerial surveys, atmospheric monitoring, and communication relays. However, due to weight and power constraints, there is a need to investigate alternate modes of propulsion to navigate in the stratosphere. Very recently, reinforcement learning…

Cited by 2 Source ↗Scholar ↗

2019

Regularization in directable environments with application to Tetris

Jan Malte Lichtenberg, Özgür Şimşek

ICML 2019oral

Learning from small data sets is difficult in the absence of specific domain knowledge. We present a regularized linear model called STEW that benefits from a generic and prevalent form of prior knowledge: feature directions. STEW shrinks weights toward each other, converging to an equal-weights sol…

Cited by 10 Source ↗PDF ↗Scholar ↗

2016

Why Most Decisions Are Easy in Tetris—And Perhaps in Other Sequential Decision Problems, As Well

Özgür Şimşek, Simón Algorta, Amit Kothiyal

ICML 2016poster

We examined the sequence of decision problems that are encountered in the game of Tetris and found that most of the problems are easy in the following sense: One can choose well among the available actions without knowing an evaluation function that scores well in the game. This is a consequence of…

Cited by 35 Source ↗PDF ↗Scholar ↗