We propose a first-order sampling method called the Metropolis-adjusted Preconditioned Langevin Algorithm for approximate sampling from a target distribution whose support is a proper convex subset of Rd. Our proposed...
详细信息
We propose a new method called the Metropolis-adjusted Mirror Langevin algorithm for approximate sampling from distributions whose support is a compact and convex set. This algorithm adds an accept-reject filter to th...
详细信息
Silicon field emitter arrays (FEAs) -based cold cathodes have shown promise in many applications under harsh environments such as x-ray sources and high-power microwave devices due to the temperature independence of t...
详细信息
The discovery of CRISPR-mediated gene activation (CRISPRa) has transformed the way in which we perform genetic screening, bioproduction and therapeutics through its ability to scale and multiplex. However, the emergen...
详细信息
This paper investigates the optimized combination of rotor and stator teeth in a three-phase switched reluctance motor featuring a connected C-core topology to attain a larger winding area and, thus, a higher electric...
详细信息
A new three-phase hybrid-excited multi-tooth switched reluctance motor with embedded permanent magnets is proposed, capable of achieving higher torque density for transportation electrification applications. Operating...
详细信息
Shape-based virtual screening is widely used in ligand-based drug design to search chemical libraries for molecules with similar 3D shapes yet novel 2D graph structures compared to known ligands. 3D deep generative mo...
详细信息
There have been several developments in renewable resources, standby sources of energy, and storage technologies. Because renewable sources are inconsistent, the best method to ensure supply continuity is to combine t...
详细信息
We study the problem of approximately transforming a sample from a source statistical model to a sample from a target statistical model without knowing the parameters of the source model, and construct several computa...
详细信息
Recent years have witnessed a paradigm shift in deep learning from task-centric model design to task-agnostic representation learning and task-specific fine-tuning. Pretrained model representations are commonly evalua...
详细信息
Recent years have witnessed a paradigm shift in deep learning from task-centric model design to task-agnostic representation learning and task-specific fine-tuning. Pretrained model representations are commonly evaluated extensively across various real-world tasks and used as a foundation for different downstream tasks. This paper proposes a solution for assessing the quality of representations in a task-agnostic way. To circumvent the need for real-world data in evaluation, we explore the use of synthetic binary classification tasks with Gaussian mixtures to probe pretrained models and compare the robustness-accuracy performance on pretrained representations with an idealized reference. Our approach offers a holistic evaluation, revealing intrinsic model capabilities and reducing the dependency on real-life data for model evaluation. Evaluated with various pretrained image models, the experimental results confirm that our task-agnostic evaluation correlates with actual linear probing performance on downstream tasks and can also guide parameter choice in robust linear probing to achieve a better robustness-accuracy trade-off. Copyright 2024 by the author(s)
暂无评论