Alert button
Picture for Jessie Rosenberg

Jessie Rosenberg

Alert button

Rapid Development of Compositional AI

Feb 12, 2023
Lee Martie, Jessie Rosenberg, Veronique Demers, Gaoyuan Zhang, Onkar Bhardwaj, John Henning, Aditya Prasad, Matt Stallone, Ja Young Lee, Lucy Yip, Damilola Adesina, Elahe Paikari, Oscar Resendiz, Sarah Shaw, David Cox

Figure 1 for Rapid Development of Compositional AI
Figure 2 for Rapid Development of Compositional AI
Figure 3 for Rapid Development of Compositional AI
Figure 4 for Rapid Development of Compositional AI

Compositional AI systems, which combine multiple artificial intelligence components together with other application components to solve a larger problem, have no known pattern of development and are often approached in a bespoke and ad hoc style. This makes development slower and harder to reuse for future applications. To support the full rapid development cycle of compositional AI applications, we have developed a novel framework called (Bee)* (written as a regular expression and pronounced as "beestar"). We illustrate how (Bee)* supports building integrated, scalable, and interactive compositional AI applications with a simplified developer experience.

* 2023 IEEE/ACM 45th International Conference on Software Engineering: New Ideas and Emerging Technologies Results Track (ICSE-NIER), Melbourne, Australia, 2023, pp. (forthcoming)  
* Accepted to ICSE 2023, NIER track 
Viaarxiv icon

SimVAE: Simulator-Assisted Training forInterpretable Generative Models

Nov 19, 2019
Akash Srivastava, Jessie Rosenberg, Dan Gutfreund, David D. Cox

Figure 1 for SimVAE: Simulator-Assisted Training forInterpretable Generative Models
Figure 2 for SimVAE: Simulator-Assisted Training forInterpretable Generative Models
Figure 3 for SimVAE: Simulator-Assisted Training forInterpretable Generative Models
Figure 4 for SimVAE: Simulator-Assisted Training forInterpretable Generative Models

This paper presents a simulator-assisted training method (SimVAE) for variational autoencoders (VAE) that leads to a disentangled and interpretable latent space. Training SimVAE is a two-step process in which first a deep generator network(decoder) is trained to approximate the simulator. During this step, the simulator acts as the data source or as a teacher network. Then an inference network (encoder)is trained to invert the decoder. As such, upon complete training, the encoder represents an approximately inverted simulator. By decoupling the training of the encoder and decoder we bypass some of the difficulties that arise in training generative models such as VAEs and generative adversarial networks (GANs). We show applications of our approach in a variety of domains such as circuit design, graphics de-rendering and other natural science problems that involve inference via simulation.

Viaarxiv icon