generative ai
Generative AI refers to algorithms, particularly those using generative models, that can create new content, such as images, music, or text, by learning from existing data, often leveraging techniques like GANs or VAEs.
- AI-Generated Video Detection via Perceptual Straightening
- Certifying Deep Network Risks and Individual Predictions with PAC-Bayes Loss via Localized Priors
- Conformal Prediction Beyond the Seen: A Missing Mass Perspective for Uncertainty Quantification in Generative Models
- Evaluating LLM-contaminated Crowdsourcing Data Without Ground Truth
- From Noise to Narrative: Tracing the Origins of Hallucinations in Transformers
- Generating Creative Chess Puzzles
- Latent Retrieval Augmented Generation of Cross-Domain Protein Binders
- More of the Same: Persistent Representational Harms Under Increased Representation
- PROFIT: A Specialized Optimizer for Deep Fine Tuning
- Puppeteer: Rig and Animate Your 3D Models
- Real-Time Hyper-Personalized Generative AI Should Be Regulated to Prevent the Rise of "Digital Heroin"
- T2SMark: Balancing Robustness and Diversity in Noise-as-Watermark for Diffusion Models
- The ML.ENERGY Benchmark: Toward Automated Inference Energy Measurement and Optimization
- Validating LLM-as-a-Judge Systems under Rating Indeterminacy
- WMCopier: Forging Invisible Watermarks on Arbitrary Images