Research Tools: code, datasets, & models

The VQA-Introspect dataset consists of 238K new perception questions which serve as sub questions corresponding to the set of perceptual tasks needed to effectively answer the complex reasoning questions in the Reasoning split of the…

The vOW4SIKE project provides C code that implements the parallel collision search algorithm by van Oorschot and Wiener (vOW). The algorithm can be instantiated for generic collision finding and for solving the supersingular isogeny problem…

This data was collected for and used in our ACL 2020 paper that demonstrates the potential to effectively combine explanations and demonstrations to learn web-based procedures. This data consists of 520 explanations and corresponding demonstrations…

SPLASH is dataset for the task of semantic parse correction with natural language feedback. The task, dataset along with baseline results are presented in: Speak to your Parser: Interactive Text-to-SQL with Natural Language Feedback Ahmed…

Using machine learning to detect beluga whale calls in hydrophone recordings. Of the five populations of beluga whales in Alaska, the Cook Inlet population is the smallest and has declined by about seventy-five percent since…

VL-BERT is a simple yet powerful pre-trainable generic representation for visual-linguistic tasks. It is pre-trained on the massive-scale caption dataset and text-only corpus, and can be fine-tuned for various down-stream visual-linguistic tasks, such as Visual…

This repository implements Ranking-Critical Training (RaCT) for Collaborative Filtering, accepted in International Conference on Learning Representations (ICLR), 2020. By using an actor-critic architecture to fine-tune a differentiable collaborative filtering model, we can improve the performance…

BERT-fused NMT is a new algorithm in which we first use BERT to extract representations for an input sequence, and then the representations are fused with each layer of the encoder and decoder of the…

KG-A2C is a reinforcement learning agent that builds a dynamic knowledge graph while exploring and generates natural language using a template-based action space – outperforming all current agents on a wide set of text-based games.

FreeLB is an adversarial training approach for improving transformer-based language models on Natural Language Understanding tasks. It accumulates the gradient in the ascent steps and updates the parameters with the accumulated gradients, which is approximately…