Decomposing a Recurrent Neural Network into Modules for Enabling Reusability and Replacement

Imtiaz, Sayem Mohammad; Batole, Fraol; Singh, Avtar; Pan, Rangeet; Cruz, Breno Dantas; Rajan, Hridesh

doi:10.1109/icse48619.2023.00093

Cited by 6 publications

(1 citation statement)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Future work includes developing techniques to repair deep neural network bugs, and exploring cases that our work was unable to detect faults (6/40) and localize errors (19/40). Recent work has also used analysis of the DNN structure to decompose it into modules[135,168]. It would be interesting to explore whether a similar mechanism can be utilized for better localization.…”

mentioning

confidence: 99%

Localizing and repairing faults in deep learning programs

Wardat

View full text Add to dashboard Cite

mentioning

confidence: 99%

Localizing and repairing faults in deep learning programs

Wardat

View full text Add to dashboard Cite

Reusing Deep Learning Models: Challenges and Directions in Software Engineering

Davis,

Jajal,

Jiang

et al. 2023

2023 IEEE John Vincent Atanasoff International Symposium on Modern Computing (JVA)

View full text Add to dashboard Cite

Inferring Data Preconditions from Deep Learning Models for Trustworthy Prediction in Deployment

Ahmed,

Gao,

Rajan

2024

Proceedings of the IEEE/ACM 46th International Conference on Software Engineering

Self Cite

View full text Add to dashboard Cite

Deep learning models are trained with certain assumptions about the data during the development stage and then used for prediction in the deployment stage. It is important to reason about the trustworthiness of the model's predictions with unseen data during deployment. Existing methods for specifying and verifying traditional software are insufficient for this task, as they cannot handle the complexity of DNN model architecture and expected outcomes. In this work, we propose a novel technique that uses rules derived from neural network computations to infer data preconditions for a DNN model to determine the trustworthiness of its predictions. Our approach, DeepInfer involves introducing a novel abstraction for a trained DNN model that enables weakest precondition reasoning using Dijkstra's Predicate Transformer Semantics. By deriving rules over the inductive type of neural network abstract representation, we can overcome the matrix dimensionality issues that arise from the backward non-linear computation from the output layer to the input layer. We utilize the weakest precondition computation using rules of each kind of activation function to compute layer-wise precondition from the given postcondition on the final output of a deep neural network. We extensively evaluated DeepInfer on 29 realworld DNN models using four different datasets collected from five different sources and demonstrated the utility, effectiveness, and performance improvement over closely related work. DeepInfer efficiently detects correct and incorrect predictions of high-accuracy models with high recall (0.98) and high F-1 score (0.84) and has significantly improved over the prior technique, SelfChecker. The average runtime overhead of DeepInfer is low, 0.22 sec for all the unseen datasets. We also compared runtime overhead using the same hardware settings and found that DeepInfer is 3.27 times faster than SelfChecker, the state-of-the-art in this area. CCS CONCEPTS• Software and its engineering → Specification languages; • Computing methodologies → Machine learning.

show abstract

Decomposing a Recurrent Neural Network into Modules for Enabling Reusability and Replacement

Cited by 6 publications

References 21 publications

Localizing and repairing faults in deep learning programs

Localizing and repairing faults in deep learning programs

Reusing Deep Learning Models: Challenges and Directions in Software Engineering

Inferring Data Preconditions from Deep Learning Models for Trustworthy Prediction in Deployment

Contact Info

Product

Resources

About