Disciplines / Research questions
Cybersecurity
10 research topics · 22 questions
Time Window for Accelerated Exploitation After Patch Release1 questions
Actual Leakage Accumulation of Differential Privacy Budget Across Multiple Queries4 questions
Monitoring-based adaptive differential-privacy defense against query-flooding model extraction at DNN scale and under colluding accounts
Attack-Parameter-to-Budget Calibration for Differentially Private Query Systems: Beyond Count Aggregates Under Repeated, Adaptive, and Auxiliary-Information Attacks
ODP filters with sublinear privacy degradation and efficient approximation of the optimal output-DP composition bound
Closing the privacy-accounting gap for linear queries in interactive differential-privacy query systems (detection, composition semantics, partial-knowledge robustness, and formal defenses against sample-amplification attacks)
Failure of Cross-language and Cross-domain Transfer of Phishing Detection Models3 questions
Generalizable phishing detection across diverse data sources and temporal shifts
Deployment-robust phishing email detection: cross-corpus generalization, prevalence-robust calibration, and elimination of corpus-artifact learning
Script generalization of language-independent phishing webpage detection: performance on non-Latin-script webpages and the mixed-alphabet failure mode
Performance Collapse of Adversarial Example Detectors Under Adaptive Attacks4 questions
Norm-agnostic robustness and architectural requirements of adversarially retrained adversarial detectors (RADAR)
Robustness of process-based (fingerprint-memory) detection of audio adversarial examples against single-step generation, fingerprint fraud, and attacks on the defense itself
Non-evadable detection of adversarial examples from internal network statistics: existence, construction, and theoretical limits
Completing the Adaptive Evaluation of Adversarial-Example Defenses: the Open Exercises of Tramèr, Carlini, Brendel and Mądry (2020)
Long-tail Effects of Dependency Graph Risk Propagation in Supply Chain Attacks1 questions
Path Sensitivity as the Root Cause of High False Positive Rates in Static Analysis2 questions
Precision, soundness, and scalability gaps in staged path-sensitive static memory-leak detection: the open questions left by SMOKE
Exact observable summaries (strongest necessary / weakest sufficient conditions) in path-sensitive static analysis: lifting the finitary-domain qualification for heap and other non-finite abstraction domains
Non-linear Boundary Between Privacy Protection and Model Utility in Federated Learning2 questions
Extending the one-local-epoch optimality of locally differentially private federated learning beyond homogeneous, i.i.d., full-participation DP-SGD
Characterizing the near-optimal utility guarantee for privacy-preserving federated learning via parameter distortion beyond the bounded-constant, small-gap, Gaussian-Bernoulli regimes
Decoupling Between Vulnerability Severity Scores and Actual Exploit Probability1 questions
Trade-off Between Security Gain and Maintenance Cost in Software Diversification Defense1 questions
Verified but Unsafe Issues Caused by Incomplete Formal Verification Specifications3 questions
Generalizing the ZBSAT security assurance framework beyond the Chinese Wall Model: verification and strict-schema-coverage testing of a Z-specified RBAC policy
Trustworthy specification autoformalization: faithfulness evaluation beyond finite test suites and extension to repository-level software
Closing the verified-but-insecure gap in the WPA3 Dragonfly/SAE handshake: extending formally verified leakage-freedom beyond primitive-level secret independence