About
Research & Development
Language Models Compete to Reveal Their Strengths and Weaknesses: An Overview of SKATE
TLDR: SKATE is a novel, automated evaluation framework where large language models (LLMs) compete by generating and solving verifiable tasks for one another. This...
Research & Development
Understanding Key Features for Better Vehicle Routing Solutions
TLDR: This research explores how machine learning and Explainable AI can identify crucial features that predict the quality of Vehicle Routing Problem (VRP) solutions....
Research & Development
A New Approach to Learning Logical Rules with Minimum Message Length
TLDR: This research introduces a Bayesian Inductive Logic Programming (ILP) method that uses Minimum Message Length (MML) to learn logical rules from noisy data....
Research & Development
Boosting Efficiency in Learning Logic Programs
TLDR: A new method for Inductive Logic Programming (ILP) introduces symmetry breaking to prune logically equivalent rules from the hypothesis space. Implemented in the...
Research & Development
AntiCheatPT: AI-Powered Cheat Detection for CS2
TLDR: A research paper introduces AntiCheatPT, a transformer-based machine learning model, and CS2CD, a public dataset, to detect cheating in Counter-Strike 2 using gameplay...
Research & Development
Explanatory AI: A New Approach to Making AI Understandable for Everyone
TLDR: A new research paradigm called "Explanatory AI" is introduced, which uses generative AI to provide human-centered, narrative-driven explanations that help users understand why...