Tech News
Midjourney Medical
Midjourney announces a medical imaging initiative aiming to reduce radiation in CT scans and make tomography cheaper and more accessible, potentially enabling earlier detection of various conditions.
GLM-5.2 is probably the most powerful text-only open weights LLM
GLM-5.2 is probably the most powerful text-only open weights LLM: chinese AI lab Z.ai released GLM-5.2 to their coding plan subscribers on June 13th, and then yesterday (June 16th) released the full open weights under an MIT license. Similar i.
Context intelligence for your data and AI agents at scale
Context intelligence for your data and AI agents at scale: agents are only as intelligent as the context they can reason over. Today, that context is scattered across data lakes, data warehouses, lakehouses, databases, and streams, and.
New research shows how AMIE, our medical AI, could help manage health conditions.
New research shows how AMIE, our medical AI, could help manage health conditions.: research in “Nature” shows our conversational AI system matches primary care physicians in complex disease management.
From the Hugging Face Hub to robot hardware with Strands Agents and LeRobot
Strands Agents and LeRobot: from the Hugging Face Hub to robot hardware with Strands Agents and LeRobot.
GitHub Repos
can1357/oh-my-pi
FallbackTypeScriptai-agentai-coding-agentmcp-use/mcp-use
FallbackTypeScriptagentic-frameworkaidtyq/magic
FallbackTypeScriptagentagiResearch Papers
RM-Bench: Benchmarking Reward Models of Language Models with Subtlety and Style
RM-Bench is a new benchmark that tests reward models on their ability to detect subtle content changes and resist style biases, revealing that even top models perform near random when style bias is present.
OpenCUA: Open Foundations for Computer-Use Agents
OpenCUA is an open-source framework for building computer-use agents, including a large dataset of human demonstrations across many apps and operating systems, and a pipeline that trains models to achieve strong performance on benchmarks.
Diversity-Aware Policy Optimization for Large Language Model Reasoning
This paper shows that promoting solution diversity during reinforcement learning training improves the reasoning performance of large language models, and proposes a practical token-level diversity objective that yields a 3.5% average improvement on math benchmarks.