Deep Learning
Embedding Trajectory for Out-of-Distribution Detection in Mathematical Reasoning Yiming Wang α Pei Zhang β, γ Baosong Y ang
Real-world data deviating from the independent and identically distributed ( i.i.d.)
Revisiting the Integration of Convolution and Attention for Vision Backbone
Convolutions (Convs) and multi-head self-attentions (MHSAs) are typically considered alternatives to each other for building vision backbones. Although some works try to integrate both, they apply the two operators simultaneously at the finest pixel granularity.
HEMM: Holistic Evaluation of Multimodal Foundation Models Paul Pu Liang
To address this need, we contribute Holistic Evaluation of Multimodal Models ( HEMM), visualized in Figure 1. HEMM, as an evaluation framework, goes beyond conventional lists of datasets to emphasize holistic benchmarking at three levels.
InternLM-XComposer2-4KHD: A Pioneering Large Vision-Language Model Handling Resolutions from 336 Pixels to 4K HD Xiaoyi Dong
The Large Vision-Language Model (L VLM) field has seen significant advancements, yet its progression has been hindered by challenges in comprehending fine-grained visual content due to limited resolution.
Accurate and Steady Inertial Pose Estimation through Sequence Structure Learning and Modulation
However, empirically, we find that, directly using the native transformer to model IMU sequences results in unacceptable jitter and inaccurate postures.