22:00

14:00

06:00

Short Introduction & Welcome to the Workshop: Ludger Paehler (Technische Universität München)

22:05

14:05

06:05

Invited Talk: Adam Paszke (Google)

Parallel-Friendly Automatic Differentiation in Dex and JAX

22:35

14:35

06:35

Invited Talk: Yuan Zhou (University of Oxford)

SYMPAIS: SYMbolic Parallel Adaptive Importance Sampling for Probabilistic Program Analysis

23:05

15:05

07:05

Contributed Talk: Uwe Naumann (RWTH Aachen University)

Differentiable Scripting

23:20

15:20

07:20

Contributed Talk: Antonio Stanziola and Simon Arridge (UCL)

A research framework for writing differentiable PDE discretizations in JAX

23:50

15:50

07:50

Invited Talk: Frank Noe (Freie Universität Berlin )

Differentiable Programming in Molecular Physics

00:20

16:20

08:20

Invited Talk: Keno Fischer (Julia Computing)

Diffractor.jl: High Level, High Performance AD for Julia

00:50

16:50

08:50

Contributed Talk: Patrick Kidger (University of Oxford)

Equinox: neural networks in JAX via callable PyTrees and filtered transformations

01:05

17:05

09:05

Contributed Talk: Deniz Bezgin (Technische Universität München)

A fully-differentiable compressible high-order computational fluid dynamics solver

01:25

17:25

09:25

Poster Session

• Extended Abstract – Enzyme.jl: Low level auto-differentiation meets high-level language
  Valentin Churavy
• GPU Accelerated Automatic Differentiation with Clad Vassil Vassilev and David Lange
• Unbiased Reparametrisation Gradient via Smoothing and Diagonalisation
  Dominik Wagner and Luke Ong
• Gradients of the Big Bang: Solving the Einstein-Boltzmann Equations with Automatic Differentiation
  James Sullivan
• Differentiable Parametric Optimization Approach to Power System Load Modeling
  Jan Drgona, Andrew August, and Elliott Skomski
• On automatic differentiation for the Matern covariance
  Oana Marin and Paul Hovland
• Neural Differentiable Predictive Control
  Jan Drgona, Aaron Tuor, and Draguna Vrabie
• AbstractDifferentiation.jl: Backend-Agnostic Differentiable Programming in Julia
  Frank Schäfer, Mohamed Tarek, Lyndon White, and Chris Rackauckas
• Aggregated type handling in AD tape implementations
  Max Sagebaum
• Backpropagation through Back substitution with a Backslash
  Ekin Akyürek, Alan Edelman, and Bernie Wang

02:45

18:45

10:45

Invited Talk: Patrick Heimbach (University of Texas at Austin)

Learning from Data through the Lens of Ocean Models, Surrogates, and their Derivatives

03:15

19:15

11:15

Invited Talk: Karen Liu (Stanford)

Learnable Physics Models

03:45

19:45

11:45

Contributed Talk: Nacime Bouziani (Imperial College London)

Escaping the abstraction: a foreign function interface for the Unified Form Language [UFL]

04:00

20:00

12:00

Contributed Talk: Alexander Lew (MIT), Mathieu Huot (University of Oxford), and Vikash Mansinghka (MIT)

Towards Denotational Semantics of AD for Higher-Order, Recursive, Probabilistic Languages

04:30

20:30

12:30

Invited Talk: Sergey Ovchinnikov (Harvard)

Differentiable Programming for Protein Sequences and Structure

05:00

21:00

13:00

Invited Talk: Harshitha Menon (Lawrence Livermore National Laboratory)

Approximate High Performance Computing Guided by Automatic Differentiation

05:30

21:30

13:30

Contributed Talk: Gordon Plotkin (University of Edinburgh)

A Complete Axiomatization of Forward Differentiation

05:45

21:45

13:45

Contributed Talk: Bhupalee Kalita (University of California, Irvine), Ryan Pederson (University of California, Irvine), Li Li (Google), Kieron Burke (University of California, Irvine)

Generalizability of density functionals learned from differentiable programming on weakly correlated spin-polarized systems

06:00

22:00

14:00

Social