Ezgi Korkmaz

6 papers · 2021–2026 · 3 conferences · across top CS/AI conferences

Achievements

🐝 Cross-Pollinator (15) 🧭 Keyword Pioneer 🌍 Conference Polyglot (3) 🏃 Academic Marathon (5) 🌉 Interdisciplinary Bridge 🗺️ Taxonomy Completionist (10) 🐺 Lone Wolf (4) 🏆 Keyword Champion (2)

Conferences

AAAI (3) ICML (2) UAI (1)

Top co-authors

Jonah Brown-Cohen (1)

Keywords

deep reinforcement learning (5) neural network (2) adversarial perturbation (2) policy robustness (2) policy learning (1) adversarial training (1) loss landscape (1) state representation (1) adversarial attack (1) decision boundary (1) feature sensitivity (1) asymptotic performance (1) adversarial feature (1) robust decision making (1) neural policy (1) fourier spectrum analysis (1) evaluation paradigm (1) empirical methodology (1) learning theory (1) adversarial robustness (1)

Papers

Principled Analysis of Deep Reinforcement Learning Evaluation and Design Paradigms AAAI 2026

Understanding and Diagnosing Deep Reinforcement Learning ICML 2024

Adversarial Robust Deep Reinforcement Learning Requires Redefining Robustness AAAI 2023

Detecting Adversarial Directions in Deep Reinforcement Learning to Make Robust Decisions ICML 2023

Deep Reinforcement Learning Policies Learn Shared Adversarial Features across MDPs AAAI 2022

Investigating vulnerabilities of deep neural policies UAI 2021