Skip to search formSkip to main contentSkip to account menu
Tensorial Mixture Models
TLDR
The effectiveness of the Tensorial Mixture Models model when tackling the problem of classification with missing data is demonstrated, leveraging TMMs unique ability of tractable marginalization which leads to optimal classifiers regardless of the missingness distribution.
Analysis and Design of Convolutional Networks via Hierarchical Tensor Decompositions
TLDR
This paper overviews a series of works written by the authors, that through an equivalence to hierarchical tensor decompositions, analyze the expressive efficiency and inductive bias of various convolutional network architectural features (depth, width, strides and more).
View PDF on arXiv
Tractable Generative Convolutional Arithmetic Circuits
TLDR
This work presents a generative model based on convolutional arithmetic circuits, a variant of Convolutional networks that computes high-dimensional functions through tensor decompositions, which theoretically achieves optimal classification and provides state of the art performance when classifying images with missing pixels.
View PDF on arXiv
Process-Level Representation of Scientific Protocols with Interactive Annotation
TLDR
Graph-prediction models are used to develop Process Execution Graphs, finding them to be good at entity identification and local relation extraction, while the corpus facilitates further exploration of challenging long-range relations.
View on ACL
A Fair Consensus Protocol for Transaction Ordering
TLDR
A quantitative measure of fairness in the protocol is defined, it is proved theoretically that fairness manipulation in Helix is significantly limited, and experiments evaluating fairness in practice are presented.
View on IEEE
Helix: A Scalable and Fair Consensus Algorithm Resistant to Ordering Manipulation
TLDR
In Helix, transactions are encrypted via a threshold encryption scheme in order to hide information from the ordering nodes, limiting censorship and front-running and to introduce a correlated sampling scheme of transactions included in a proposed block.
View Paper
Boosting Dilated Convolutional Networks with Mixed Tensor Decompositions
TLDR
It is proved that interconnecting dilated convolutional networks can lead to expressive efficiency, and it is shown that even a single connection between intermediate layers can already lead to an almost quadratic gap, which in large-scale settings typically makes the difference between a model that is practical and one that is not.
View PDF on arXiv
Language (Re)modelling: Towards Embodied Language Understanding
TLDR
It is argued that the use of grounding by metaphoric reasoning and simulation will greatly benefit NLU systems, and a system architecture along with a roadmap towards realizing this vision is proposed.
View on ACL
Playing by the Book: An Interactive Game Approach for Action Graph Extraction from Text
TLDR
A novel approach to action-graph extraction from materials science papers is proposed, Text2Quest, where procedural text is interpreted as instructions for an interactive game, which can complement existing approaches and enables richer forms of learning compared to static texts.
View on ACL
Scaling Creative Inspiration with Fine-Grained Functional Facets of Product Ideas
TLDR
This work proposes a novel computational representation that automatically breaks up products into fine-grained functional facets, and designs similarity metrics that support granular matching between functional facets across ideas, and uses them to build a novel functional search capability that enables expressive queries for mechanisms and purposes.
...
...
By clicking accept or continuing to use the site, you agree to the terms outlined in our Privacy Policy, Terms of Service, and Dataset License