GPT4o-Receipt Dataset for AI Document Forensics Study
Explore GPT4o-Receipt, a dataset and study revealing AI vs human accuracy in detecting AI-generated financial documents and receipts.
TikZilla: Advanced Text-to-TikZ with RL & Quality Data
Discover TikZilla, a cutting-edge Text-to-TikZ model using high-quality data and reinforcement learning to generate accurate scientific figures.
Agentified Framework for Evaluating Logical Reasoning AI
Discover a robust agentified framework for reproducible and auditable assessment of logical reasoning agents, boosting AI evaluation accuracy and reliabili...
CIRCLE Framework: Real-World AI Evaluation Guide
Discover the CIRCLE framework for evaluating AI in real-world contexts, enhancing accountability, decision-making, and governance in AI deployment.
CollectiveKV: Efficient KV Sharing for Fast Sequential Rec
CollectiveKV reduces KV cache size by 99% in sequential recommendation, cutting latency and storage without losing performance.