About
Research & Development
Unpacking Opinion Dynamics in Large Language Models: Interaction vs. Inherent Biases
TLDR: Researchers developed a Bayesian framework to quantify how interaction and three biases (topic, agreement, anchoring) influence opinion changes in LLM discussions. They found...
Research & Development
AxelSMOTE: A New Agent-Based Approach to Balancing Imbalanced Datasets
TLDR: AxelSMOTE is a novel agent-based oversampling algorithm inspired by Axelrod's cultural dissemination model, designed to address class imbalance in machine learning. It overcomes...
Research & Development
A New Method for Reliable AI Out-of-Distribution Detection in Noisy Environments
TLDR: A new research paper introduces NOODLE, a framework to improve Out-of-Distribution (OOD) detection in AI systems, especially when training data contains noisy labels....
Research & Development
LLMs Advance Insider Threat Detection Through Synthetic Data Generation
TLDR: A new study introduces an ethically grounded approach using large language models (LLMs) like Claude Sonnet 3.7 and GPT-4o to synthesize and detect...
Research & Development
Unveiling the Inner Workings: How AI Models Fabricate Information from Noise
TLDR: This research paper investigates how and when hallucinations arise in transformer AI models. Using Sparse Autoencoders (SAEs), the study reveals that transformers impose...
Research & Development
Enhancing Image Generation with Text-Based Thinking and Refinement
TLDR: The paper introduces Interleaving Reasoning Generation (IRG), a framework that improves text-to-image generation by alternating between text-based thinking and image synthesis. It first...