cross-domain generalization
Cross-domain generalization is the capability of a model to perform well on tasks in different domains from those it was trained on. This ability is critical for models applied in diverse and dynamic real-world environments.
- DeltaFlow: An Efficient Multi-frame Scene Flow Estimation Method
- Drag-and-Drop LLMs: Zero-Shot Prompt-to-Weights
- GUI-Rise: Structured Reasoning and History Summarization for GUI Navigation
- Latent Retrieval Augmented Generation of Cross-Domain Protein Binders
- MAT-Agent: Adaptive Multi-Agent Training Optimization
- MiNT: Multi-Network Transfer Benchmark for Temporal Graph Learning
- Mitigating Forgetting in LLM Fine-Tuning via Low-Perplexity Token Learning
- No Object Is an Island: Enhancing 3D Semantic Segmentation Generalization with Diffusion Models
- Revisiting Reinforcement Learning for LLM Reasoning from A Cross-Domain Perspective
- SSTAG: Structure-Aware Self-Supervised Learning Method for Text-Attributed Graphs
- Time Series Generation Under Data Scarcity: A Unified Generative Modeling Approach
- Two Experts Are All You Need for Steering Thinking: Reinforcing Cognitive Effort in MoE Reasoning Models Without Additional Training