Empowering Diffusion Models on the Embedding Space for Text Generation.
Zhujin GaoJunliang GuoXu TanYongxin ZhuFang ZhangJiang BianLinli XuPublished in: NAACL-HLT (2024)
Keyphrases
- text generation
- diffusion models
- embedding space
- euclidean space
- low dimensional
- diffusion model
- natural language generation
- manifold learning
- dimensionality reduction
- information diffusion
- graph embedding
- high dimensional
- input space
- geometric structure
- data points
- natural language
- social networks
- nonlinear dimensionality reduction
- geodesic distance
- theorem prover
- viral marketing
- influence maximization
- high dimensional data
- euclidean distance
- greedy algorithm
- semi supervised
- shape analysis
- principal component analysis
- natural language processing
- social media
- feature space
- network structure
- supervised learning
- pattern recognition
- computer vision