zero-shot
A capability of AI models to generalize and perform tasks without prior explicit training on the specific task, often relying on transferred knowledge from similar tasks.
- Evaluating Generalization Capabilities of LLM-Based Agents in Mixed-Motive Scenarios Using Concordia
- From Play to Replay: Composed Video Retrieval for Temporally Fine-Grained Videos
- Memory-Enhanced Neural Solvers for Routing Problems
- RESAnything: Attribute Prompting for Arbitrary Referring Segmentation
- SplashNet: Split‑and‑Share Encoders for Accurate and Efficient Typing with Surface Electromyography