blog | news | updates
winter 2022
| MolT5 - our work evaluating transformers for text-to-text transfer between natural language and molecular information accepted for oral presentation at EMNLP2022. | [perma] |
summer 2022
| blog | Patents filed for our effort to deploy facsimiles of real biological connectomes as interchangeable reservoirs in artificial neural networks. |
| blog | We shutdown our investigation of a wearable lab and micro-needle array for online proteome analysis. Some patents from the effort apear here. |
spring 2022
| blog | We filed a ton of patents for technology supporting our work on advanced diagnostics and monitoring for mental health including filings on the earliest known use of transformer architectures for electrophysiological representation learning. |
summer 2021
| blog | We open-source our work enabling the use of adaptable reservoir layers for building neural networks in Keras. BYOR - Bring your own reservoir! |
spring 2021
| Our paper on bVAEs and disentanglement for intepretability and classification of clinical factors from EEG presented at ICLR 2021. | [perma] |
winter 2021
| blog | I joined the Cortex team at X. We do math, science, and machine learning for zero-to-one projects. |
winter 2020
| Spotlight talk accepted at the NeuRIPS 2021 Machine Learning for Mobile Health workshop: Using Convolutional Variational Autoencoders to Predict Post-Trauma Health Outcomes from Actigraphy Data. | [perma] | |
| Paper on the correspondence between N400 amplitude and similarity processing accepted for publication at Neuropsychologia. | [perma] |
fall 2020
| blog | My Main project at X was opensourced today: Sharing Project Amber with the mental health community. |
fall 2019
| Our framework for combating online disinformation accepted to the Comparative Approaches to Disinformation workshop hosted by the Berkman Klien Center for Internet and Society. | [perma] |
summer 2019
| Submitted our latest on semantic knowledge organization (preprint at PsyArXiv) that provides evidence that N400 amplitude reliably differs for processing of taxonomic catgeory members and thematic associates; critically, the amplitude delta varies consistently with people's offline similarity judgment behavior. | [perma] | |
| Our paper introducing the category status hypothesis for knowledge transfer accepted at JEP:LMC. We show in a series of experiments that learning a novel relational category through category construction improves spontaneous transfer. | [perma] | |
| blog | Accepted a research scientist role focused on computational neursoscience and ml with a project in stealth mode housed at X (Alphabet's moonshot factory). |
spring 2019
| Paper accepted for presentation at ICML '19 AI for Social Good Workshop. The project tested an ensemble CNN-LSTM pipeline for text classification. | [perma] | |
| blog | Short thread with resources on digitization, government records, and publishing decisions of the Mueller Report by the U.S. DOJ |
winter 2019
| Colleagues and I submitted SIMON (preprint on arXiv). SIMON is a semantic classifier that uses the features of a character-level convolutional neural network to predict class membership. The algorithm shows impressive performance on tasks as diverse as social media account user age, tabular data column types (e.g., GPS coordinates, ordinal values, postal codes, etc.), and Spam classification. | [perma] | |
| Paper published at Cognition: Similarity is as similarity does? A critical inquiry into the effect of thematic association on similarity. This work provides a better understanding of the curious case of thematic similarity, where thematic integration-driven similarity judgment behavior is found to be less frequent than thought. | [perma] |
fall 2018
| Paper published at the Journal of Research in Science Teaching - category construction in the classroom as a technique to reduce the use of scientifically-invalid (i.e., intuitive) beliefs about inheritance | [perma] |
summer 2018
| The first behavioral work from my dissertation is under review (preprint at PsyArXiv) wherein we discovered a surprising inconsistency in human similarity judgements across the timecourse of a single experimental session | [perma] | |
| code | Colleagues and I (mostly them!) have open-sourced IBEX, a topic extraction library in python | [perma] |
| code | Published UNICORN (UNsupervised Image Clustering with Object Recognition Network), a python library that classifies images based on their similarity with respect to the activation pattern produced by the Inception V3 network trained on Imagenet. | [perma] |
| Paper on automatic dataset topic summarization accepted for presentation at the ICML '18 AutoML workshop | [perma] |
spring 2018
| code | Colleagues and I published APE (Abstractive Prediction via Embedding), an open-source PIP-installable library that predicts more abstract labels (i.e. higher-order category membership) for a provided set of concepts | [perma] |
| Symposium on relational category learning accepted at CogSci 2018 | [perma] | |
| code | Published an open-source PIP-installable library that performs object recognition and OCR | [perma] |
winter 2017
| blog | Experimenting with text generation to create novel Ferengi Rules of Acquisition |
fall 2017
| I successfully defended my dissertation, a behavioral and electrophysiological investigation of human similarity judgments. | [perma] | |
| Submitted the first preprint from the LaRC Lab - Category Construction Promotes Transfer | [perma] |
summer 2017
| blog | Quick tutorial on grid search with Catlearn and Catlearn Supplementals |
| blog | Brief introduction to machine learning and cognitive modeling with Catlearn and Catlearn Supplementals |
spring 2017
| code | Nolan Conaway and I wrote a suite of helper functions to assist with modeling in the Catlearn ecosystem | [perma] |
| code | I wrote a DIVA (DIVergent Autoencoder) model for the Catlearn R library | [perma] |
summer 2016
| I gave a talk at CogSci 2016 on learning categories via feature switching. Here's the paper. | [perma] |
summer 2015
| Presented a paper on teaching for transfer in middle school science at the American Educational Research Association annual meeting | [perma] |
fall 2014
| Paper on analogical learning at the Chicago Children's Museum accepted for publication at Cognitive Science | [perma] |
summer 2012
| blog | Safest Bike Route from Binghamton (city) to Binghamton (university) |