搜索“Agent”

iles navigation 👋 join us on Discord and WeChat Introduction We introduce Intern-S2-397B, our most capable multimodal foundation model for scientific intelligence and long-horizon agents. Intern-S2-397B scales along three critical dimensions: pre-training, reinforcement-learning task coverage, and interactive agent environments. By combining a new vision-language pre-training paradigm with large-scale multi-task reinforcement learning and long-horizon agent reinforcement learning, Intern-S2-397B

匹配字段正文

he next frontier in AI, demanding powerful world models to realize its full potential. World models should reconstruct, generate, and simulate 3D worlds; and allow both humans and agents to interact with them. Spatially intelligent world models will transform a wide variety of industries over the coming years. Two months ago we shared a preview of Marble, our World Model that creates 3D worlds from image or text prompts. Since then, Marble has been available to an early set of beta users to crea

匹配字段正文

Meta has launched Muse Wonder and Muse Wonder Full, two AI models built for personal and professional agents, with capabilities spanning long-term memory, multi-step task execution, and coding. The models are part of Meta’s effort to scale its AI efforts. Meta trained both models were trained at the company’s gigawatt-scale Prometheus data centre. The company also plans to use the model to develop specialised models for wider access through Muse. The Muse Wonder series combines two capabilities

匹配字段正文

penAI models. Model targets specific decisions Decision-1 supports yes-or-no judgments, choices among predefined options, ratings, and rubric-based assessments of AI responses and agent actions, Microsoft said. Developers can use its outputs to classify inputs, rank work, verify responses, or determine the next action in an AI agent workflow. Unlike a general-purpose LLM, which can generate explanations and other natural-language content, decision models return probabilities for the available an

匹配字段正文

x) at 48. The gap between Spark 1.3 and Wonder releases was only 6 weeks. Opus 5.5 leads at 58 (@TheRundownAI); Wonder Full improved on all ten index evals. The largest gains were agentic:Terminal-Bench 4.0: 33% → 57%; GDPval-AA v2.1: 1681 → 1857 Elo; AA-Briefcase: 1581 → 1729 Elo; Where Wonder Full ranks: top three on 6 of 10 evals. 1 on GDP.pdf (33%) and on CritPt (33%); 2 on GDPval; 3 on AA-Briefcase, AutomationBench-AA (71%) and SciCode (63%); Wonder (non-Full) gains: Terminal-Bench 33% → 52

匹配字段正文

Kopparapu in a video accompanying the launch. The new models are the first frontier model release since Scale AI founder Alexandr Wang restructured Meta’s AI efforts in August of last year . Since then, Meta has released its lightweight Muse Spark model and Muse agent but has avoided the kind of state-of-the-art model releases OpenAI and Anthropic are known for. According to the post, the models were trained at Meta’s Prometheus data center in central Ohio , which came online earlier this year.

匹配字段正文

between OpenAI and Anthropic so far. The climb has been impressive, but the scarier part for rivals is everything around the model — the data centers, billions of users, and a hit agent app in Muse that are hard qualities for other labs to replicate. TOGETHER WITH BOX 💭 Reimagine enterprise AI at BoxWorks 2026 The Rundown: Unlocking enterprise AI begins by connecting agents to where your business knowledge lives: your content. Box securely connects unstructured data to the tools your teams use d

匹配字段正文

gh an API and through Muse Code, a tool for writing software with AI. Meta says the best version of the model will power features across its platforms, including its Muse personal agent. Zuckerberg overhauled his company’s AI research last year by sinking billions into the recently formed Meta Superintelligence Labs, and hiring the former Scale AI CEO Alexandr Wang to lead it. This effort has sometimes been rocky, with some researchers not sticking around for long. Meta was once a trailblazer in

匹配字段正文

Meta is making up for lost time in the AI race. After launching Muse Spark , the first model from its revamped Superintelligence Labs division, and buzzy AI agent Muse , it’s now launching two new models it claims can go toe-to-toe with the industry’s best. The company says its new models can rival leading models from OpenAI and Anthropic, including GPT-6 Astra, Claude Opus 5.5, and Claude Fable 5.1, pointing to results from internal testing on several industry benchmarks. It’s unclear how they

匹配字段正文

and well-rounded performance across a range of benchmarks. Inkling-Small Inkling Compared to Inkling, Inkling-Small achieves comparable performance with much less compute. Across agentic tool use (Terminal-Bench 2.1), reasoning (HLE text-only, no tools), and instruction following (IFBench) benchmarks, Inkling-Small is more efficient than Inkling and competitive with other models in its weight class. Furthermore, its variable thinking effort lets users easily adapt it to target their use case, b

匹配字段正文

Capabilities Generalist model; Agentic coding and tool use One-shot web app with embedded browser use Design Arena; Cohesively styled artifacts; Multiplayer game created through long refinement loop; Controllable thinking effort; Multimodality; Epistemics; Safety; Benchmarking Inkling; The making of Inkling Architecture; Training; RL at scale; Inkling-Small; Customizing Inkling Inkling availability Our mission is to build AI that extends human will and judgment. We have developed a platform that

匹配字段正文

on settings; Fork1 (1); Star15 (15)You must be signed in to star a repository Repository files navigation README LLaDA-UI is an MoE-based, block-wise diffusion vision-language GUI agent. It combines a native dynamic-resolution vision encoder with the LLaDA2.0-mini-base diffusion language backbone, and generates reasoning and GUI actions through the same block-wise diffusion decoder. Figure 1. LLaDA-UI GUI-agent performance and qualitative block-wise diffusion decoding. The radar chart and GUI ob

匹配字段正文

ficationsYou must be signed in to change notification settings; Fork87 (87); Star1.1k (1.1k)You must be signed in to star a repository UI-Venus is a general-purpose foundation GUI agent for mobile apps, web platforms, and desktop operating systems using only screenshots as input. Repository files navigation README English | 简体中文 A general-purpose foundation GUI agent for mobile apps, web platforms, and desktop operating systems — scaling environments, tasks, and feedback jointly within one close

匹配字段正文

ing and inference overhead. RoboBrain-X0 delivers state-of-the-art performance in embodied reasoning tasks, laying a robust foundation for developing versatile, real-world robotic agents and advancing embodied intelligence research. 2025-09-29: 🔥 RoboBrain-X0-preview was released 📆 Todo Release model checkpoint for RoboBrain-X0-Preview based on Fast Tokenizer; Release quick inference example for RoboBrain-X0; Release training and evaluation codes for RoboBrain-X0; Release omnisat-based RoboBrain

匹配字段正文

nd Muse Wonder Full, our most capable models to date, built on the groundwork we laid with Muse Spark. The models deliver state-of-the-art performance on personal and professional agentic tasks. Built for Real World Needs Whether the job is navigating an insurance claim, understanding a lease, or making sense of someone’s finances, a model has to interpret the information in front of it and work through the problem. Muse Wonder and Muse Wonder Full bring that capability to expert domains includi

匹配字段正文

decision-making, and retrosynthesis. Table 1 3.2 Genomics and quantitative biology To evaluate GPT-Rosalind-5.5 on genomics understanding and quantitative biology, we designed an agentic evaluation on long horizon, end-to-end genomics analysis tasks. This evaluation assesses agentic performance on long-horizon quantitative tasks: based on realistic scientific data, can an agent plan valid analysis, QC, modeling, and corrections to arrive at decision-relative answers? Genebench spans a variety o

匹配字段正文

d our gpt-oss usage policy. Developed with feedback from the open-source community, these text-only models are compatible with our Responses API and are designed to be used within agentic workflows with strong instruction following, tool use like web search and Python code execution, and reasoning capabilities—including the ability to adjust the reasoning effort for tasks that don’t require complex reasoning. The models are customizable, provide full chain-of-thought (CoT), and support Structure

匹配字段正文

propriate safeguards, and improving how we measure and report model performance for younger users. 3.2 Vision We ran the image input evaluations originally introduced with ChatGPT Agent. These tests use disallowed requests that combine text and images and measure whether the model’s responses are classified as safe. GPT‑6 Sol (October) and GPT‑6 Luna (October) performed on-par with GPT-5.6 Sol (August) and GPT-5.6 Luna (August), respectively, except for a regression on the extremism vision evalu

匹配字段正文

des performance information for GPT-6 Astra on the new, updated versions; September 29, 2026: We added an Appendix: dots with information about dots, remarkably capable, always-on agents built to handle everything. Dots are built on existing model and agent capabilities, powered by GPT-6 Astra 1. Safety Overview: GPT-6 Astra Today, we are releasing GPT-6 Astra, the most capable model we have ever broadly deployed. Astra is our first model to reach the Critical level of cybersecurity capability u

匹配字段正文

e for users on Meta's first-party API, while Muse Wonder Full is currently only available for Muse Code users. Meta notes that each model is designed for personal and professional agent use Key takeaways: ➤ Both models exhibit a material jump in overall Intelligence Index score. Muse Wonder Full (max) improves on Muse Spark 1.3 (max) across all ten Intelligence Index evaluations, gaining 8 points to 56. The largest gains were seen in agentic evaluations: Terminal-Bench 4.0 (+24 points, 33% to 57

匹配字段正文