Efficient Error Certification for Physics-Informed Neural Networks
May 17, 2023
| | | | |
Computer Science
Physics
Mathematics
Machine Learning
Mathematical Physics
Mathematical Physics
Recent work provides promising evidence that Physics-Informed Neural Networks
(PINN) can efficiently solve partial differential equations (PDE). However,
previous works have failed to provide guarantees on the worst-case residual
error of a...
Photo-induced charge-transfer renormalization in NiO
May 17, 2023
| | | | | | | | | | and others
Physics
Strongly Correlated Electrons
Photo-doped states in strongly correlated charge transfer insulators are
characterized by d-d and d-p interactions and the resulting intertwined
dynamics of charge excitations and local multiplets. Here we use femtosecond
x-ray abso...
Reprompting: Automated Chain-of-Thought Prompt Inference Through Gibbs
Sampling
May 17, 2023
| |
Computer Science
Machine Learning
Artificial Intelligence
Computation and Language
We introduce Reprompting, an iterative sampling algorithm that automatically
learns the Chain-of-Thought (CoT) recipes for a given task without human
intervention. Through Gibbs sampling, Reprompting infers the CoT recipes that
work consist...
Real-Time Flying Object Detection with YOLOv8
May 17, 2023
| | |
Computer Science
Computer Vision and Pattern Recognition
Machine Learning
This paper presents a generalized model for real-time detection of flying
objects that can be used for transfer learning and further research, as well as
a refined model that achieves state-of-the-art results for flying object
detection. We...
ACRoBat: Optimizing Auto-batching of Dynamic Deep Learning at Compile
Time
May 17, 2023
| | |
Computer Science
Machine Learning
Dynamic control flow is an important technique often used to design
expressive and efficient deep learning computations for applications such as
text parsing, machine translation, exiting early out of deep models and so on.
The control flow...
Characterization of exchangeable measure-valued P\'olya urn sequences
May 17, 2023
|
Mathematics
Probability
Measure-valued P\'olya urn sequences (MVPS) are a generalization of the
observation processes generated by k-color P\'olya urn models, where the
space of colors X is a complete separable metric space and the urn
composition is ...
Effect of the atomic structure of complexions on the active
disconnection mode during shear-coupled grain boundary motion
May 17, 2023
| | |
Physics
Materials Science
The migration of grain boundaries leads to grain growth in polycrystals and
is one mechanism of grain-boundary-mediated plasticity, especially in
nanocrystalline metals. This migration is due to the movement of
dislocation-like defects, cal...
A Simple and Elementary Proof of Zorn's Lemma
May 17, 2023
Mathematics
Logic
Combinatorics
Zorn's Lemma is a well-known equivalent of the Axiom of Choice. It is usually
regarded as a topic in axiomatic set theory, and its historically standard
proof (from the Axiom of Choice) relies on transfinite recursion, a
non-elementary set-...
Chiral limit and origin of topological flat bands in twisted transition
metal dichalcogenide homobilayers
May 17, 2023
| |
Physics
Mesoscale and Nanoscale Physics
Strongly Correlated Electrons
The observation of zero field fractional quantum Hall analogs in twisted
transition metal dichalcogenides (TMDs) asks for a deeper understanding of what
mechanisms lead to topological flat bands in two-dimensional heterostructures,
and what...
Adaptive aggregation of Monte Carlo augmented decomposed filters for
efficient group-equivariant convolutional neural network
May 17, 2023
| | | | | |
Computer Science
Computer Vision and Pattern Recognition
Group-equivariant convolutional neural networks (G-CNN) heavily rely on
parameter sharing to increase CNN's data efficiency and performance. However,
the parameter-sharing strategy greatly increases the computational burden for
each added p...