embedding space
A lower-dimensional vector space where data points (such as words or images) are represented as vectors. This space captures semantic relationships and structural properties, enabling effective comparisons and operations on the embedded entities.
- Adaptive and Multi-scale Affinity Alignment for Hierarchical Contrastive Learning
- Adv-SSL: Adversarial Self-Supervised Representation Learning with Theoretical Guarantees
- Attention Sinks: A 'Catch, Tag, Release' Mechanism for Embeddings
- Autoencoding Random Forests
- Cost-Aware Contrastive Routing for LLMs
- Diversity Is All You Need for Contrastive Learning: Spectral Bounds on Gradient Magnitudes
- Embeddings as Probabilistic Equivalence in Logic Programs
- Enhancing Compositional Reasoning in CLIP via Reconstruction and Alignment of Text Descriptions
- Failure Prediction at Runtime for Generative Robot Policies
- Geometric Imbalance in Semi-Supervised Node Classification
- GnnXemplar: Exemplars to Explanations - Natural Language Rules for Global GNN Interpretability
- Hierarchical Retrieval: The Geometry and a Pretrain-Finetune Recipe
- MetaDefense: Defending Fine-tuning based Jailbreak Attack Before and During Generation
- Refusal Direction is Universal Across Safety-Aligned Languages
- Semantic-guided Diverse Decoding for Large Language Model
- ToF-IP: Time-of-Flight Enhanced Sparse Inertial Poser for Real-time Human Motion Capture
- Topology-aware Graph Diffusion Model with Persistent Homology
- UniHG: A Large-scale Universal Heterogeneous Graph Dataset and Benchmark for Representation Learning and Cross-Domain Transferring
- Visual Diversity and Region-aware Prompt Learning for Zero-shot HOI Detection