Picture for Alexandre de Brébisson

Alexandre de Brébisson

A Cheap Linear Attention Mechanism with Fast Lookups and Fixed-Size Representations

Add code
Sep 19, 2016
Figure 1 for A Cheap Linear Attention Mechanism with Fast Lookups and Fixed-Size Representations
Figure 2 for A Cheap Linear Attention Mechanism with Fast Lookups and Fixed-Size Representations
Viaarxiv icon

Exact gradient updates in time independent of output size for the spherical loss family

Add code
Jun 26, 2016
Figure 1 for Exact gradient updates in time independent of output size for the spherical loss family
Figure 2 for Exact gradient updates in time independent of output size for the spherical loss family
Figure 3 for Exact gradient updates in time independent of output size for the spherical loss family
Figure 4 for Exact gradient updates in time independent of output size for the spherical loss family
Viaarxiv icon

The Z-loss: a shift and scale invariant classification loss belonging to the Spherical Family

Add code
May 27, 2016
Figure 1 for The Z-loss: a shift and scale invariant classification loss belonging to the Spherical Family
Figure 2 for The Z-loss: a shift and scale invariant classification loss belonging to the Spherical Family
Figure 3 for The Z-loss: a shift and scale invariant classification loss belonging to the Spherical Family
Figure 4 for The Z-loss: a shift and scale invariant classification loss belonging to the Spherical Family
Viaarxiv icon

Theano: A Python framework for fast computation of mathematical expressions

Add code
May 09, 2016
Figure 1 for Theano: A Python framework for fast computation of mathematical expressions
Figure 2 for Theano: A Python framework for fast computation of mathematical expressions
Figure 3 for Theano: A Python framework for fast computation of mathematical expressions
Figure 4 for Theano: A Python framework for fast computation of mathematical expressions
Viaarxiv icon

An Exploration of Softmax Alternatives Belonging to the Spherical Loss Family

Add code
Feb 28, 2016
Figure 1 for An Exploration of Softmax Alternatives Belonging to the Spherical Loss Family
Figure 2 for An Exploration of Softmax Alternatives Belonging to the Spherical Loss Family
Figure 3 for An Exploration of Softmax Alternatives Belonging to the Spherical Loss Family
Figure 4 for An Exploration of Softmax Alternatives Belonging to the Spherical Loss Family
Viaarxiv icon

Artificial Neural Networks Applied to Taxi Destination Prediction

Add code
Sep 21, 2015
Figure 1 for Artificial Neural Networks Applied to Taxi Destination Prediction
Figure 2 for Artificial Neural Networks Applied to Taxi Destination Prediction
Figure 3 for Artificial Neural Networks Applied to Taxi Destination Prediction
Figure 4 for Artificial Neural Networks Applied to Taxi Destination Prediction
Viaarxiv icon

Efficient Exact Gradient Update for training Deep Networks with Very Large Sparse Targets

Add code
Jul 14, 2015
Figure 1 for Efficient Exact Gradient Update for training Deep Networks with Very Large Sparse Targets
Figure 2 for Efficient Exact Gradient Update for training Deep Networks with Very Large Sparse Targets
Figure 3 for Efficient Exact Gradient Update for training Deep Networks with Very Large Sparse Targets
Viaarxiv icon