Optimism Induction Attack on Deep Reinforcement Learning with Control Barrier Function Safety Filter for Autonomous Driving
Bibliographic record
Abstract
As autonomous vehicles (AVs) increasingly incorporate Reinforcement Learning (RL) into their decision-making processes, ensuring both security and stability becomes paramount. This paper introduces the Optimism Induction Attack (OIA), a novel adversarial strategy specifically targeting Deep Reinforcement Learning (DRL) agents. In contrast to traditional adversarial attacks—such as the Fast Gradient Sign Method (FGSM), which primarily degrades overall performance—OIA strategically exploits the agent’s misperception of state safety, causing it to overestimate safety margins and consequently make suboptimal decisions in critical scenarios. While OIA is broadly applicable to any actor-critic RL algorithm, we conduct a case study on a Proximal Policy Optimization (PPO) -trained Adaptive Cruise Control (ACC) agent protected by Control Barrier Function (CBF). Our analysis examines system performance using metrics such as collision rate, jerk, and engine torque. The results reveal that OIA significantly undermines both safety and efficiency, emerging as a subtler and more effective adversarial threat than FGSM, as evidenced by increased collision rates despite the nominal safety guarantees provided by CBFs. This work advances the field of adversarial machine learning in AVs by highlighting the urgent need for more robust defense mechanisms capable of countering sophisticated attacks like OIA, particularly in safety-critical applications.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".