# Dataset Articoli collegati

Il Centro Notizie HTX fornisce gli articoli più recenti e le analisi più approfondite su "Dataset", coprendo tendenze di mercato, aggiornamenti sui progetti, sviluppi tecnologici e politiche normative nel settore crypto.

First Long-Horizon Doc2Repo Training Dataset: Code Agents Move Beyond Bug Fixing and Begin Creating Repositories

With the advancement of LLM Code Agents, the research focus is shifting towards long-horizon, real-world tasks, moving beyond simple bug fixes to full repository generation. To address this, researchers from Renmin University of China introduced the DeNovoSWE dataset. This dataset focuses on long-term software engineering tasks, specifically the "document-to-repository" challenge—generating an entire, executable code repository from a task description. The DeNovoSWE construction method employs a Divide & Conquer approach. It breaks down target repositories into core capabilities and uses a multi-agent Draft-Critic-Repair workflow to automatically generate high-quality, evaluation-aligned task documents. The dataset also implements difficulty-aware filtering to balance quality and diversity. The result is a high-quality, anti-leakage dataset of 4,818 instances. Experiments show that models trained on DeNovoSWE achieve significant improvements in long-horizon repository generation. For instance, Qwen3-30B-A3B-Instruct's performance on the BeyondSWE-Doc2Repo benchmark increased from 5.8% to 47.2%, and on NL2RepoBench from 4.3% to 23.0%. Similar gains were observed with stronger backbones, demonstrating that dedicated long-horizon training data is crucial for advancing Code Agents from maintainers to architects capable of planning and building complete software projects from scratch.

marsbit06/25 08:51

First Long-Horizon Doc2Repo Training Dataset: Code Agents Move Beyond Bug Fixing and Begin Creating Repositories

marsbit06/25 08:51

For the First Time, Pure Human Video Pretrained VLA for Dexterous Manipulation: Deployable with Minimal Fine-Tuning Data

For the first time, a purely human-video-pretrained Vision-Language-Action (VLA) model for dexterous manipulation requires only a small amount of data for fine-tuning to achieve successful real-world deployment. Achieving human-level dexterous manipulation remains a core challenge in robotics. While multi-fingered hands offer hardware potential, Visual-Language-Action (VLA) models lag behind due to the high cost of collecting diverse, high-quality robot data. A novel framework, VITRA, developed by Microsoft Research Asia and Tsinghua University, addresses this by automatically transforming massive, unlabeled real-world human activity videos into a structured V-L-A training dataset. Key innovations include precise 3D hand motion annotation from monocular video, atomic action segmentation based on hand-speed minima, and automated instruction generation using VLMs combined with 3D trajectory visualization. This process created a massive dataset of 1 million clips. Pretrained exclusively on this human video data, the VLA model (combining a VLM backbone with a Diffusion Transformer action expert) demonstrates strong zero-shot hand motion prediction in unseen environments. Crucially, it requires minimal fine-tuning (~1.2k demonstrations) on real robot data to achieve high-success-rate dexterous manipulation tasks like grasping, placing, pouring, and sweeping on hardware like the Realman robot with the XHAND1 dexterous hand. The model shows exceptional generalization to novel objects and environments. The research also observes promising scaling behavior, where performance improves with more pretraining data, paving the way for more generalized embodied intelligence.

marsbit06/08 08:54

For the First Time, Pure Human Video Pretrained VLA for Dexterous Manipulation: Deployable with Minimal Fine-Tuning Data

marsbit06/08 08:54

Embodied Intelligence Breakthrough: Amap Fully Open-Sources Universal Robot Base Model ABot-M0

Embodied Intelligence Breakthrough: AutoNavi Open-Sources Universal Robot Base Model ABot-M0 AutoNavi has announced the full open-source release of ABot-M0, the world's first unified architecture-based embodied manipulation base model. This model is designed to enable "one general brain to adapt to multiple forms of robots," aiming to break down barriers between heterogeneous hardware and accelerate the adoption of embodied intelligence in industrial and household settings. ABot-M0 demonstrated exceptional performance in industry tests, achieving a task success rate of 80.5% on the Libero-Plus benchmark—a nearly 30% improvement over the previous benchmark, Pi0. It also set new state-of-the-art records on benchmarks like Libero and RoboCasa. The open-source release addresses long-standing challenges in the field, such as data isolation and deployment difficulties, by providing resources across three key dimensions: - **Data:** The UniACT dataset, the largest of its kind, with over 6 million real operation trajectories and full data pipeline tools. - **Algorithm:** The model architecture and training framework, featuring innovative components like Action Manifold Learning (AML) and a dual-stream perception architecture. - **Model:** End-to-end pre-trained models and a complete toolchain for out-of-the-box deployment, significantly lowering the barrier to adaptation. According to AutoNavi's ABot-M0 technical lead, this open-source initiative aims to build a bridge between academic research and industrial application, enabling robots of various forms to possess a smart, reliable, and universal "brain."

marsbit04/01 08:19

Embodied Intelligence Breakthrough: Amap Fully Open-Sources Universal Robot Base Model ABot-M0

marsbit04/01 08:19

AI2 Releases Fully Open-Source Web Agent MolmoWeb: Controlling Web Pages Using Only "Vision"

AI2 has released MolmoWeb, a groundbreaking, fully open-source web agent that operates solely by analyzing screenshots, marking a significant leap in vision-driven web navigation. Unlike traditional agents that rely on DOM, MolmoWeb captures and interprets visual data to make decisions—such as clicking, scrolling, or typing—making its process transparent and robust. Despite its compact size (4B and 8B parameters), MolmoWeb performs impressively: it scores 78.2% on the WebVoyager benchmark, nearing OpenAI’s proprietary o3 model (79.3%), and achieves up to 94.7% success with multiple attempts. It even surpasses Anthropic’s Claude3.7 in UI element localization. AI2 also released MolmoWebMix, a massive open dataset with 36K human-browsing tasks, over 2.2M screenshot-QA pairs, and GPT-4o-verified synthetic data. The model and data are fully available on Hugging Face and GitHub under Apache 2.0, promoting transparency and collaboration in AI development. Challenges remain in complex instructions, logins, and legal compliance.

marsbit03/26 01:39

AI2 Releases Fully Open-Source Web Agent MolmoWeb: Controlling Web Pages Using Only "Vision"