About
Research & Development
Anecdoctoring: A New Method for Global AI Disinformation Red-Teaming
TLDR: Anecdoctoring is a novel red-teaming approach that automatically generates adversarial prompts across diverse languages and cultures to evaluate AI models for disinformation risks....
Research & Development
YAC: A New AI System for Exploring Biomedical Data with Natural Language and Interactive Visuals
TLDR: YAC (Yet Another Chatbot) is a novel system that integrates natural language processing with interactive data visualizations to enhance biomedical data discovery. Developed...
Research & Development
Boosting Language Model Reasoning with Reinforcement Learning on Pre-Training Data
TLDR: Reinforcement Learning on Pre-Training data (RLPT) is a new method for optimizing large language models (LLMs) that uses reinforcement learning directly on vast...
Research & Development
Advancing Human Motion Understanding with Adversarially-Refined VQ-GANs
TLDR: This research introduces an Adversarially-Refined VQ-GAN framework with dense motion tokenization to compress spatio-temporal heatmaps of human motion. The method effectively eliminates reconstruction...
Research & Development
SloPalSpeech: Unlocking Advanced Speech Recognition for Slovak with Parliamentary Data
TLDR: Researchers have introduced SloPalSpeech, a new 2,800-hour dataset of Slovak speech derived from parliamentary proceedings. This large-scale, meticulously processed dataset has been used...
Research & Development
Large Language Models Reshaping Operations Research: A Comprehensive Overview
TLDR: This research paper surveys the integration of Large Language Models (LLMs) into Operations Research (OR), categorizing methods into automatic modeling, auxiliary optimization, and...