All published AI news and analysis.

OpenAI is increasing its engagement with developers, businesses, and communities in Brazil. This expansion aims to foster AI adoption throughout the country.

Researchers introduce UltraPIPS, a library of Learned Perceptual Image Patch Similarity (LPIPS) metrics specifically designed for B-mode ultrasound images. The library utilizes foundation models trained on ultrasound data, demonstrating improved correlation with downstream task performance compared to models trained on natural or general medical images.

This article details the process of training and fine-tuning multi-vector embedding models using the Sentence Transformers library. It explores techniques for creating more sophisticated embedding representations that can capture nuanced semantic relationships.

OpenAI has released a report detailing how students and educators leverage ChatGPT for continuous learning. The findings highlight its role in providing support that extends beyond traditional classroom settings.

Researchers have introduced CorporateBench (CB), a new Q&A benchmark designed to evaluate LLMs on enterprise-scale document collections. CB features human-validated, multi-task Q&A across corpora exceeding 230,000 documents, simulating corporate communication networks.

DeepMind has launched the world's first double-blind evaluation for a proprietary AI model, Gemini Flash Lite. This method uses cryptographic technology to ensure neither the model nor the evaluators can see each other's sensitive data, preventing benchmark contamination.

Researchers have introduced VBVR-Pro, a closed-loop testbed designed to make native visual reasoning trainable and verifiable. The suite includes 300 procedurally generated tasks and verifiable reward scorers, aiming to overcome limitations in scalable training and reliable feedback for visual reasoning models.

A new technique called Quantization-Aware Healing allows for the creation of compressed 4-bit AI models that achieve superior performance compared to their original full-precision versions. This method focuses on optimizing the quantization process to maintain and even enhance model accuracy.

OpenAI has released version 0.149.1 of its Codex CLI tool. This update is part of ongoing maintenance and improvements to the command-line interface for interacting with Codex models.

Travel company loveholidays is leveraging OpenAI Codex to democratize software development across its business. This initiative aims to empower various teams to translate their ideas into functional products more rapidly.