Location History:
- Cambridge, MA (US) (2022)
- Albany, NY (US) (2022)
Company Filing History:
Years Active: 2022
Title: Halley Fede: Innovator in Code Representation and Semantic Similarity
Introduction
Halley Fede is an accomplished inventor based in Cambridge, MA (US). He has made significant contributions to the field of computer science, particularly in the area of code representation and error detection. With a total of 2 patents, Fede's work is at the forefront of technological innovation.
Latest Patents
Fede's latest patents include groundbreaking methods for representing source code in vector space to detect errors. This computer-implemented method involves parsing source code into an abstract syntax tree, which is then traversed to produce a sequence of tokens. Token embeddings are constructed for a subset of these tokens and inputted into an encoder artificial neural network. The decoder is then run to learn the inputted token embeddings, ultimately using the final internal cell state of the encoder as the code representation vector for error detection.
Another notable patent focuses on generating vector representations of code that capture semantic similarity. This method involves collecting two training sets of code snippets, one representing semantic similarity and the other representing dissimilarity. A machine learning model generates vector representations of these code snippets, and the model is trained to minimize the loss value, effectively capturing the semantic similarity in the code.
Career Highlights
Halley Fede is currently employed at International Business Machines Corporation (IBM), where he continues to innovate and develop new technologies. His work has garnered attention for its practical applications in software development and error detection.
Collaborations
Fede collaborates with talented individuals such as David Wehr and Eleanor Pence, contributing to a dynamic and innovative work environment.
Conclusion
Halley Fede's contributions to the field of computer science through his patents and work at IBM highlight his role as a leading inventor in code representation and semantic similarity. His innovative methods are paving the way for advancements in error detection and software development.