Skip to main content

Turing Honors Scholars Program

Exploring Annotator Rationales for Active Learning with Transformers

Filtering data in transformers

12/14/2022 - For decades, natural language processing (NLP) has provided methods for computers to understand language in a way that mimics humans. Since they are built on transformers, complex neural network layers, these large language models' decision making processes are usually incomprehensible to humans and require large amounts of data to be trained properly. In the past, researchers have tried to remedy this by having models explain their decisions by providing rationales, short excerpts of data that contributed most to the label.

All That Jazz: Improving Automated Piano Note Transcription

hands playing a piano

11/17/2021 - Any fan of jazz music can attest to the beauty of musical improvisation. However, many famous improvisational piano pieces aren't recorded in sheet music. “There's a lot of music that exists in the world that doesn't have musical transcriptions because it was played improvisationally—virtuosos that never decided to write anything down,” explained Varun Rajaram. This is because transcribing the notes of a piece (especially polyphonic pieces where multiple notes play at a time) is a difficult task even for skilled musicians.

Undergraduate Experience Prepares Alumnus to Launch Startup

Denis Ignatovich (right) and Grant Passmore, co-founders of Imandra

05/26/2020 - On August 1, 2012, the global financial services firm Knight Capital, which was at the time the largest trader in U.S. equities, lost $460 million due to a “technology breakdown.” One of their trading servers housed defective code, causing the group irreparable damage. Almost exactly a year later, a Goldman Sachs computer glitch resulted in a number of erroneous trades, resulting in a loss of over $100 million for the company.