ICML 2025 Past Generative models
Data in Generative Models - The Bad, the Ugly, and the Greats
DIG-BUG
- Submission deadline
- May 29, 2025, 12:00 UTC imported from OpenReview — check the website for extensions
- Submission portal
- OpenReview
- Notes
- Auto-imported from the OpenReview venue record on 2026-06-10 — please verify and enrich (topics are keyword-guessed).
Accepted papers (37)
Fetched from OpenReview (v2) on 2026-06-10.
-
A Data-Centric Safety Framework for Generative Models: Adversarial Fingerprint Detection and Attribution
-
A Representation Engineering Perspective on the Effectiveness of Multi-Turn Jailbreaks
-
Backdooring VLMs via Concept-Driven Triggers
-
Both Text and Images Leaked! A Systematic Analysis of Data Contamination in Multimodal LLMs
-
Cascading Adversarial Bias from Injection to Distillation in Language Models
-
COREVQA: A Crowd Observation and Reasoning Entailment Visual Question Answering Benchmark
-
Data Cartography for Detecting Memorization Hotspots and Guiding Data Interventions in Generative Models
-
Detective SAM: Adapting SAM to Localize Diffusion-based Forgeries via Embedding Artifacts
-
Diversity Boosts AI-Generated Text Detection
-
DP-AdamW: Investigating Decoupled Weight Decay and Bias Correction in Private Deep Learning
-
FaceSafe: An Inpainting Pipeline for Privacy-Compliant Scalable Image Datasets
-
Firm Foundations for Membership Inference Attacks Against Large Language Models
-
Generalizing Trust: Weak-to-Strong Trustworthiness in Language Models
-
Ghost in the Cloud: Your Geo-Distributed Large Language Models Training is Easily Manipulated
-
Implementing Adaptations for Vision AutoRegressive Model
-
Improvement-Guided Iterative DPO for Diffusion Models
-
In-Context Bias Propagation in LLM-Based Tabular Data Generation
-
JailbreakLoRA: Your Downloaded LoRA from Sharing Platforms might be Unsafe
-
Layer-wise Influence Tracing: Data-Centric Mitigation of Memorization in Diffusion Models
-
Lookahead Bias in Pretrained Language Models
-
MAD-MAX: Modular And Diverse Malicious Attack MiXtures for Automated LLM Red Teaming
-
Model-based Large Language Model Customization as Service
-
Optimal Defenses Against Data Reconstruction Attacks
-
Optimization and Robustness-Informed Membership Inference Attacks for LLMs
-
OVERT: A Benchmark for Over-Refusal Evaluation on Text-to-Image Models
-
Preference Leakage: A Contamination Problem in LLM-as-a-judge
-
R&B: Breaking the Data Mixing Bottleneck with Just 0.01% Overhead
-
Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy
-
RN-F: A Novel Approach for Mitigating Contaminated Data in Large Language Models
-
SOSBENCH: Benchmarking Safety Alignment on Scientific Knowledge
-
Spectral Manifold Harmonization for Graph Imbalanced Regression
-
Training Diffusion Models with Noisy Data via SFBD Flow
-
TruthLens: Training-Free Data Verification for Deepfake Images via VQA-style Probing
-
Unlocking Post-hoc Dataset Inference with Synthetic Data
-
Watermarking Image Autoregressive Models
-
Weak-to-strong Generalization via Formative Learning from Student Demonstrations & Teacher Evaluation
-
Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets