📑Alibaba: "당신 손 대신, AI가 폰·PC를 직접 조작합니다"

이 호에서 다룬 논문

• RTX

  Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents

  Alibaba (Qwen) · Foundation GUI Agent, Cross-platform Automation, Online RL

• AI RESEARCH

  AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis

  New York University · Claim-Centered Retrieval, Provenance Grounding, Chemistry Literature

• WORLD MODEL

  PhiZero: A World Model Built Around Physical Language

  CASIA (중국과학원 자동화연구소) · World Model, Physical Language, Reason-then-Render

• AGENTS AI

  Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering

  Independent (OpenMLE / Frontis AI) · Recursive Self-Improvement, AI4AI, ML Engineering Agent

• GENERATIVE

  Chimera: Designing and Chinchilla-Scaling Hybrid Visual Diffusion Transformers

  Adobe · Hybrid Diffusion Transformer, Linear Attention, Compute-optimal Scaling

• EFFICIENCY

  Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory

  Shanghai Jiao Tong University · Parametric Memory, Long-term Memory Module, Parameter Efficiency

• VISION

  Beacon: Knowing When and How to Perform Agentic Visual Reasoning

  Kuaishou · Agentic Visual Reasoning, Mode Adaptiveness, Adaptive Tool Use

• ROBOTICS

  ACE-Data-0: Human-Centric Ambient Capture as Embodied Data Engine

  Nanyang Technological University · Embodied Data Engine, Multisensory Capture, Imitation Learning

• EFFICIENCY

  MindForge: Teaching Small Language Models Whole-Life-Cycle Software Engineering via Source-Free Program Synthesis

  Queen's University (SAIL Lab) · Program Synthesis from Scratch, Source-Free Environment, Small Model Fine-tuning

• EFFICIENCY

  β-OPSD: Deriving with Policy Optimization, Training with Self-Distillation

  University of Maryland · On-policy Self-Distillation, Policy Optimization, KL Regularization