performance gaps
Performance gaps refer to discrepancies in model accuracy or effectiveness across diverse data distributions, user demographics, or environments, highlighting areas for improvement in AI system design.
- Can Large Multimodal Models Understand Agricultural Scenes? Benchmarking with AgroMind
- ColorBench: Can VLMs See and Understand the Colorful World? A Comprehensive Benchmark for Color Perception, Reasoning, and Robustness
- Embodied Web Agents: Bridging Physical-Digital Realms for Integrated Agent Intelligence
- Measure-Theoretic Anti-Causal Representation Learning
- SEC-bench: Automated Benchmarking of LLM Agents on Real-World Software Security Tasks
- Seeking and Updating with Live Visual Knowledge
- Strategic Classification with Non-Linear Classifiers