UBA-ORL: Unlearning-Activated Backdoor Attacks on Offline Reinforcement Learning

arXiv:2609.22711 · cs.CR · Submitted 2026-09-19 · Read on arXiv

cs.CR

Submitted: 2026-09-19

Updated: 2026-09-19

Comments: Accepted at IEEE ICDM 2026. arXiv version: 11 pages, 6 figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Reinforcement Learning — 3× in this paper · explained in 4 episodes
• Unlearning-activated Backdoor Attack on Offline Reinforcement Learning — 1× in this paper

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs