Organization
IBM
5 stories · sorted newest first · 📡 RSS
IBM and UC Berkeley Study Reveals Failure Signatures in Large Language Models
New research using MAST identifies why LLMs like Gemini-3-Flash, Kimi-K2, and GPT-OSS-120B fail in real-world IT automation tasks.
IBM Launches Compact Vision-Language Model for Enterprise Document Understanding
The new IBM Granite 4.0 3B Vision model excels in specialized extraction tasks and is designed to run without heavy hardware, maki
IBM Introduces VAKRA: A Comprehensive Benchmark for AI Agents in Enterprise Settings
VAKRA, a new tool-grounded executable benchmark from IBM Research, evaluates AI agents' ability to reason end-to-end in complex, m
IBM's Granite 4.1: High-Quality, Open-Source LLM for Adult Industry
IBM's Granite 4.1 models offer strong performance in various tasks and are designed to address limitations of larger mixture-of-ex
IBM's Granite Embedding Multilingual R2 Revolutionizes Open-Source Search with 32K Context Window
New open-source multilingual embedding models from IBM promise to simplify search and retrieval across multiple languages, handlin