performance gap
The difference in performance between two models or systems, often highlighted in studies assessing the effectiveness of AI methods. This can indicate areas where improvements are needed or where one approach significantly outperforms another.
- A Smooth Sea Never Made a Skilled SAILOR: Robust Imitation via Learning to Search
- ACT as Human: Multimodal Large Language Model Data Annotation with Critical Thinking
- AuroRA: Breaking Low-Rank Bottleneck of LoRA with Nonlinear Mapping
- CAPability: A Comprehensive Visual Caption Benchmark for Evaluating Both Correctness and Thoroughness
- ChartMuseum: Testing Visual Reasoning Capabilities of Large Vision-Language Models
- ExAct: A Video-Language Benchmark for Expert Action Analysis
- KeyDiff: Key Similarity-Based KV Cache Eviction for Long-Context LLM Inference in Resource-Constrained Environments
- Large Language Models for Lossless Image Compression: Next-Pixel Prediction in Language Space is All You Need
- Mixing Expert Knowledge: Bring Human Thoughts Back To the Game of Go
- ReinAD: Towards Real-world Industrial Anomaly Detection with a Comprehensive Contrastive Dataset
- Same Task, Different Circuits: Disentangling Modality-Specific Mechanisms in VLMs
- Time Series Generation Under Data Scarcity: A Unified Generative Modeling Approach
- Understanding and Improving Fast Adversarial Training against $l_0$ Bounded Perturbations
- Vanish into Thin Air: Cross-prompt Universal Adversarial Attacks for SAM2
- macOSWorld: A Multilingual Interactive Benchmark for GUI Agents