Zero-Run auditing supplies valid lower bounds on differential privacy parameters from fixed member and non-member datasets by modeling and correcting distribution-shift confounding via causal-inference techniques.
Extracting training data from large language models
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 8roles
background 1polarities
background 1representative citing papers
A²utoLPBench is a generator that produces unlimited LP word problems with ground-truth answers known by construction via inverse-KKT, bundled with a Docker environment for agent evaluation.
Infilling extraction on diffusion language models extracts up to three times more verbatim sequences than prefix methods and achieves higher recall on redacted emails than autoregressive models.
DAPRO adaptively spends a fixed interaction budget across multi-turn LLM conversations and builds valid lower bounds on time-to-jailbreak, with coverage error scaling as the square root of the mean censoring weight.
MADreMIA amplifies membership inference signals by showing that memorized samples maintain higher coherence and slower degradation in chained regeneration trajectories than non-members.
Vision-language models exhibit perceptual fragility and fail to consistently respect privacy constraints when operating in simulated physical environments, with performance declining in cluttered scenes and under conflicting commands.
Neuro-symbolic RAG framework with formal ontology achieves 78.6-point recall improvement on AI-generated phishing threats while keeping precision above 98% and false positive rate at 0.16% under privacy constraints.
Personal agents require edge deployment to preserve high-fidelity local context and zero-latency loops, as claimed through three structural shifts away from cloud-centric designs.
citing papers explorer
-
Privacy Auditing with Zero (0) Training Run
Zero-Run auditing supplies valid lower bounds on differential privacy parameters from fixed member and non-member datasets by modeling and correcting distribution-shift confounding via causal-inference techniques.
-
A$^{2}$utoLPBench: An Auto-Generated, Agent-Friendly LP Benchmark via Inverse-KKT Construction
A²utoLPBench is a generator that produces unlimited LP word problems with ground-truth answers known by construction via inverse-KKT, bundled with a Docker environment for agent evaluation.
-
Extracting Training Data from Diffusion Language Models via Infilling
Infilling extraction on diffusion language models extracts up to three times more verbatim sequences than prefix methods and achieves higher recall on redacted emails than autoregressive models.
-
How Many Iterations to Jailbreak? Dynamic Budget Allocation for Multi-Turn LLM Evaluation
DAPRO adaptively spends a fixed interaction budget across multi-turn LLM conversations and builds valid lower bounds on time-to-jailbreak, with coverage error scaling as the square root of the mean censoring weight.
-
Amplifying Membership Signal Through Chained Regeneration
MADreMIA amplifies membership inference signals by showing that memorized samples maintain higher coherence and slower degradation in chained regeneration trajectories than non-members.
-
How Far Are VLMs from Privacy Awareness in the Physical World? An Empirical Study
Vision-language models exhibit perceptual fragility and fail to consistently respect privacy constraints when operating in simulated physical environments, with performance declining in cluttered scenes and under conflicting commands.
-
CyberCane: Neuro-Symbolic RAG for Privacy-Preserving Phishing Detection with Formal Ontology Reasoning
Neuro-symbolic RAG framework with formal ontology achieves 78.6-point recall improvement on AI-generated phishing threats while keeping precision above 98% and false positive rate at 0.16% under privacy constraints.
-
Beyond Scaling: Agents Are Heading to the Edge
Personal agents require edge deployment to preserve high-fidelity local context and zero-latency loops, as claimed through three structural shifts away from cloud-centric designs.