Every Ling 2.0 model uses the same sparse Mixture of Experts layer. Each layer has 256 routed experts and one shared expert. The router picks 8 routed experts for every token, the shared expert is ...

How do you convert real agent traces into reinforcement learning RL transitions to improve policy LLMs without changing your existing agent stack? Microsoft AI team releases Agent Lightning to help ...

Can an open source MoE truly power agentic coding workflows at a fraction of flagship model costs while sustaining long-horizon tool use across MCP, shell, browser, retrieval, and code? MiniMax team ...

What is new in Granite 4.0 Nano series? Granite 4.0 Nano consists of four model lines and their base counterparts. Granite 4.0 H 1B uses a hybrid SSM based architecture and is about 1.5B parameters.

Can we render long texts as images and use a VLM to achieve 3–4× token compression, preserving accuracy while scaling a 128K context toward 1M-token workloads? A team of researchers from Zhipu AI ...

AI companies use model specifications to define target behaviors during training and evaluation. Do current specs state the intended behaviors with enough precision, and do frontier models exhibit ...

Web agents often fail when layouts shift or when tasks require long sequences. WALT targets this failure mode by mining site functionality offline, then exposing it as tools that encapsulate ...

In this article we will analyze how Google, OpenAI, and Anthropic are productizing ‘agentic’ capabilities across computer-use control, tool/function calling, orchestration, governance, and enterprise ...

Orchestration Host routes across many servers/tools App-local chaining Agent/toolkit routes intents → operations ...

Computer-use agents have been limited to primitives. They click, they type, they scroll. Long action chains amplify grounding errors and waste steps. Apple Researchers introduce UltraCUA, a foundation ...

The post emphasizes coding parity with Sonnet 4 and computer-use gains relative to Sonnet 4 under these scaffolds. Users should replicate with their own orchestration, tool stacks, and thinking ...

What’s in the release? SKUs and variants: The new additions comprise four dense models— Qwen3-VL-4B and Qwen3-VL-8B, each in Instruct and Thinking editions—alongside FP8 versions of the 4B/8B Instruct ...