Martin Gubri
9 papers · 2022–2026 · 6 venues · across top CS/AI venues
Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.
Achievements
Jump to papers ↓+1 more ↓ Show less ↑
🌍 Conference Polyglot · Silver 🌉 Interdisciplinary Bridge · Bronze 🗺️ Taxonomy Completionist (17) 🐣 Hot Topic Early Bird 🧭 Keyword Pioneer
🔥
Unstoppable · Bronze
Top co-authors
Research topics
Artificial Intelligence
(8)
Core AI
(7)
Machine Learning
(5)
Application Areas
(3)
Learning Types · Machine Learning
(3)
Deep Learning
(2)
Natural Language Processing
(2)
Security & Privacy
(2)
Optimization & Theory
(2)
Privacy
(2)
Resources & Methods
(2)
Bayesian & Probabilistic
(1)
Learning Types · Deep Learning
(1)
Models
(1)
Keywords
large language model
(5)
membership inference
(2)
privacy leakage
(2)
posterior distribution
(1)
model identification
(1)
data privacy
(1)
black-box attack
(1)
bayesian neural network
(1)
adversarial example
(1)
adversarial prompt
(1)
black-box detection
(1)
training data
(1)
privacy vulnerability
(1)
reasoning trace
(1)
prompt injection
(1)
black-box model
(1)
membership inference attack
(1)
confidence estimation
(1)
personally identifiable information
(1)
white-box model
(1)
Papers
MASEval: Extending Multi-Agent Evaluation from Models to Systems
ACL 2026
Privacy Collapse: Benign Fine-Tuning Can Break Contextual Privacy in Language Models
ACL 2026
Leaky Thoughts: Large Reasoning Models Are Not Private Thinkers
EMNLP 2025
Scaling Up Membership Inference: When and How Attacks Succeed on Large Language Models
NAACL 2025
Calibrating Large Language Models Using Their Generations Only
ACL 2024
TRAP: Targeted Random Adversarial Prompt Honeypot for Black-Box Identification
ACL 2024
ProPILE: Probing Privacy Leakage in Large Language Models
NeurIPS 2023
Efficient and transferable adversarial examples from bayesian neural networks
UAI 2022
LGV: Boosting Adversarial Example Transferability from Large Geometric Vicinity
ECCV 2022