전체 글 썸네일형 리스트형 Less is More: Recursive Reasoning with Tiny Networks https://arxiv.org/abs/2510.04871?ref=refetch.io Less is More: Recursive Reasoning with Tiny NetworksHierarchical Reasoning Model (HRM) is a novel approach using two small neural networks recursing at different frequencies. This biologically inspired method beats Large Language models (LLMs) on hard puzzle tasks such as Sudoku, Maze, and ARC-AGI while traarxiv.org 초록(Abstract)Hierarchical Reasoni.. 더보기 Vision-Zero: Scalable VLM Self-Improvement via Strategic Gamified Self-Play https://arxiv.org/html/2509.25541v1 Vision-Zero: Scalable VLM Self-Improvement via Strategic Gamified Self-PlayVision-Zero: Scalable VLM Self-Improvement via Strategic Gamified Self-Play Qinsi Wang1, Bo Liu2, Tianyi Zhou3, Jing Shi4, Yueqian Lin1, Yiran Chen1 Hai Helen Li1, Kun Wan4, Wentian Zhao4* Corresponding authors.arxiv.org 초록(Abstract)비전-언어 모델(Vision–Language Models, VLMs.. 더보기 Kaggle 마스터 달성 더보기 Introducing Gemini 2.5 Flash Image, our state-of-the-art image model https://developers.googleblog.com/en/introducing-gemini-2-5-flash-image/?utm_source=pytorchkr&ref=pytorchkr Introducing Gemini 2.5 Flash Image, our state-of-the-art image model- Google Developers BlogToday, we’re excited to introduce Gemini 2.5 Flash Image (aka nano-banana), our state-of-the-art image generation and editing model. This update enables you to blend multiple images into a single im.. 더보기 Polynomial Composition Activations: Unleashing the Dynamics of Large Language Models https://arxiv.org/abs/2411.03884 Polynomial Composition Activations: Unleashing the Dynamics of Large Language ModelsTransformers have found extensive applications across various domains due to the powerful fitting capabilities. This success can be partially attributed to their inherent nonlinearity. Thus, in addition to the ReLU function employed in the original transfoarxiv.org 초록(Abstract)Tra.. 더보기 kaiber.biz https://kaiber.biz/nne-py_en/ KAIBER NN Editor for PyTorch_enRather than supporting a variety of frameworks in a generic fashion, PyTorch specialization focuses on a single design methodology, simplifies integration with previously developed Python code, and reduces bugs by automatically checking for consistency witkaiber.biz 더보기 asanAI https://github.com/NormanTUD/asanAI?utm_source=chatgpt.com 더보기 DINOv3 https://arxiv.org/abs/2508.10104 DINOv3Self-supervised learning holds the promise of eliminating the need for manual data annotation, enabling models to scale effortlessly to massive datasets and larger architectures. By not being tailored to specific tasks or domains, this training paradigm haarxiv.org 초록자기지도학습(self-supervised learning)은 수작업 데이터 라벨링의 필요성을 제거하고, 모델이 거대한 데이터셋과 대규모 아키텍처로 손쉽게 확장할 수.. 더보기 이전 1 2 3 4 ··· 72 다음
