Christopher Malon NEC Labs America Christopher Malon is a Senior Researcher in the Machine Learning Department at NEC Laboratories America. He received his Bachelor’s degree in Mathematics from the University of Chicago and holds a PhD in Mathematics from the Massachusetts Institute of Technology.

He focuses on building more trustworthy AI systems, leading projects in fact verification, opinion summarization, and hallucination avoidance in generative AI. In AI for healthcare, he contributed to NEC’s research and development of the world’s first automatic cancer screening system based on digital biopsy images, and now leads efforts at NEC on trustworthy medical report generation for multiple modalities. A key theme in his fact verification and medical report generation research is recognizing missing or incomplete information, to pinpoint which details need additional consideration by an AI agent or a human.

With a strong track record in interdisciplinary collaboration, Dr. Malon’s research is helping NEC translate machine learning innovations into trustworthy AI partners to improve productivity in healthcare and other domains.

Posts

Team Papelo: Transformer Networks at FEVER

/in /by

We develop a system for the FEVER fact extraction and verification challenge that uses a high precision entailment classifier based on transformer networks pretrained with language modeling, to classify a broad set of potential evidence. The precision of the entailment classifier allows us to enhance recall by considering every statement from several articles to decide upon each claim. We include not only the articles best matching the claim text by TFIDF score, but read additional articles whose titles match named entities and capitalized expressions occurring in the claim text. The entailment module evaluates potential evidence one statement at a time, together with the title of the page the evidence came from (providing a hint about possible pronoun antecedents). In preliminary evaluation, the system achieves .5736 FEVER score, .6108 label accuracy, and .6485 evidence F1 on the FEVER shared task test set.

Read more