OpenAI has launched a new family of models called GPT-4.1. They focus on coding, and are exclusively available through the company's API.

The models constitute IBM’s most multimodal release yet.

LLMs are evolving into a landscape of smaller, purpose-built models.

Learn about Gecko, a rubric-based and interpretable autorater for evaluating generative AI models.

MIT researchers found that vision-language models, widely used to analyze medical images, do not understand negation words like “no” and “not.” This could cause them to fail unexpectedly when asked to…

Due to its significant resource demands and requirement for rapid scalability, AI is the ultimate hybrid application.

Google and Cohere now offer takeout versions of their proprietary frontier LLMs, which users can run in their own data centers.

My weekly AI governance curation to help you stay ahead | Edition #221

We discovered why language models catastrophically fail on long conversations: when old tokens are removed to save memory, models produce complete gibberish. We found models dump massive attention onto the first few tokens as "attention sinks"—places to park…

Popular small language models are reshaping the AI landscape by combining efficiency, strong reasoning, factual accuracy, tool use, and broad accessibility.