generative tasks
Generative tasks in AI involve producing new data or samples that resemble a given data distribution. Examples include generating images, text, or audio and often utilize techniques in generative modeling such as GANs or VAEs.
- CCS: Controllable and Constrained Sampling with Diffusion Models via Initial Noise Perturbation
- Compress & Cache: Vision token compression for efficient generation and retrieval
- Compressed and Smooth Latent Space for Text Diffusion Modeling
- Curriculum Model Merging: Harmonizing Chemical LLMs for Enhanced Cross-Task Generalization
- Dense Associative Memory with Epanechnikov Energy
- GeoAda: Efficiently Finetune Geometric Diffusion Models with Equivariant Adapters
- Multi-Token Prediction Needs Registers
- Self Iterative Label Refinement via Robust Unlabeled Learning
- Training a Scientific Reasoning Model for Chemistry
- Why Diffusion Models Don’t Memorize: The Role of Implicit Dynamical Regularization in Training
- Why Diffusion Models Don’t Memorize: The Role of Implicit Dynamical Regularization in Training