Doc2X-优快云博客

原创 Microsoft COCO: Common Objects in Context 翻译

https://arxiv.org/pdf/1405.0312Microsoft COCO: Common Objects in Context 翻译Microsoft COCO：上下文中的常见对象

2024-11-22 17:41:51 656

原创 Concept Arithmetics for Circumventing Concept Inhibition in Diffusion Models 翻译

https://arxiv.org/pdf/2404.13706Concept Arithmetics for Circumventing Concept Inhibition in Diffusion Models翻译概念算术：绕过扩散模型中的概念抑制

2024-11-22 17:32:41 1049

原创 Rasterized Edge Gradients: Handling Discontinuities Differentiably 翻译

https://arxiv.org/pdf/2405.02508Rasterized Edge Gradients: Handling Discontinuities Differentiably 翻译光栅化边缘梯度：以可微分方式处理不连续性

2024-11-22 17:20:11 619

原创 Minimalist Vision with Freeform Pixels 翻译

https://cave.cs.columbia.edu/Statics/publications/pdfs/Klotz_ECCV24.pdfMinimalist Vision with Freeform Pixels 翻译极简主义视觉与自由形式像素

2024-11-22 17:12:36 771

原创 Do Efficient Transformers Really Save Computation? 翻译

https://arxiv.org/pdf/2402.13934Do Efficient Transformers Really Save Computation? 翻译高效的Transformer真的节省计算吗

2024-11-22 16:59:52 863

原创 RNNS ARE NOT TRANSFORMERS (YET): The Key Bottleneck on In-context Retrieval 翻译

https://arxiv.org/pdf/2402.18510RNNS ARE NOT TRANSFORMERS (YET): The Key Bottleneck on In-context Retrieval 翻译RNNS 还不是 TRANSFORMERS（目前）：上下文检索的关键瓶颈

2024-11-22 16:56:04 923

原创 HydraLoRA: An Asymmetric LoRA Architecture for Efficient Fine-Tuning 翻译

https://arxiv.org/pdf/2404.19245HydraLoRA: An Asymmetric LoRA Architecture for Efficient Fine-Tuning 翻译HydraLoRA: 一种用于高效微调的不对称LoRA架构

2024-11-22 16:47:03 1268

原创 VideoAgent: Self-Improving Video Generation 翻译

https://arxiv.org/pdf/2410.10076VideoAgent: Self-Improving Video Generation 翻译VideoAgent: 自我改进的视频生成

2024-11-22 16:34:13 1201

原创 Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation 翻译

https://arxiv.org/pdf/2410.13848Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation 翻译Janus：解耦视觉编码以实现统一的多模态理解和生成

2024-11-22 16:26:20 907

原创 Beyond Preferences in AI Alignment 翻译

https://arxiv.org/pdf/2408.16984Beyond Preferences in AI Alignment 翻译超越偏好在AI对齐中的应用

2024-11-22 16:17:45 715

原创 Model Swarms: Collaborative Search to Adapt LLM EXPERTS VIA SWARM INTELLIGENCE 翻译

https://arxiv.org/pdf/2410.11163Model Swarms: Collaborative Search to Adapt LLM EXPERTS VIA SWARM INTELLIGENCE 翻译模型集群：通过集群智能协作搜索来适应大型语言模型专家

2024-11-22 13:49:12 784

原创 WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language Models 翻译

https://arxiv.org/pdf/2405.14768WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language ModelsWISE：重新思考终身模型编辑大型语言模型的知识记忆

2024-11-22 12:46:16 800

原创 OKAMI: Teaching Humanoid Robots Manipulation Skills through Single Video Imitation 翻译

https://arxiv.org/pdf/2410.11792OKAMI: Teaching Humanoid Robots Manipulation Skills through Single Video ImitationOKAMI：通过单个视频模仿教人形机器人操作技能

2024-11-22 11:58:01 908

原创 DuoAttention: EFFICIENT LONG-CONTEXT LLM INFERENCE WITH RETRIEVAL AND STREAMING HEADS 翻译

https://arxiv.org/pdf/2410.10819DuoAttention: EFFICIENT LONG-CONTEXT LLM INFERENCE WITH RETRIEVAL AND STREAMING HEADSDuoAttention: 带有检索和流式头的高效长上下文LLM推理

2024-11-22 11:57:37 720

原创 VERIFIED: A Video Corpus Moment Retrieval Benchmark for Fine-Grained Video Understanding 翻译

https://arxiv.org/pdf/2410.08593VERIFIED: A Video Corpus Moment Retrieval Benchmark for Fine-Grained Video Understanding验证：用于细粒度视频理解的帧检索基准

2024-11-21 16:11:04 725

原创 Computational Bottlenecks of Training Small-scale Large Language Models 翻译

https://arxiv.org/pdf/2410.19456Computational Bottlenecks of Training Small-scale Large Language Models小规模大型语言模型训练的计算瓶颈

2024-11-21 16:03:17 1076

原创 PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic Manipulation 翻译

https://arxiv.org/pdf/2410.10394PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic ManipulationPIVOT-R: 用于机器人操作的原语驱动航点感知世界模型

2024-11-21 15:51:44 962

原创 THE GEOMETRY OF CONCEPTS: Sparse Autoencoder Feature Structure 翻译

https://arxiv.org/pdf/2410.19750THE GEOMETRY OF CONCEPTS: Sparse Autoencoder Feature Structure概念的几何：稀疏自编码器特征结构

2024-11-21 15:37:20 371

原创 Reward Centering 翻译

https://arxiv.org/pdf/2405.09999Reward Centering奖励中心化

2024-11-21 14:56:02 970

原创 AFLOW: Automating Agentic Workflow Generation 翻译

https://arxiv.org/pdf/2410.10762AFLOW: Automating Agentic Workflow GenerationAFLOW: 自动化代理工作流生成

2024-11-21 14:41:58 875

原创 FAN: Fourier Analysis Networks 翻译

FAN: Fourier Analysis Networks 翻译原文链接：https://arxiv.org/pdf/2410.02675

2024-11-21 14:22:01 2443

原创 CryoSTAR: Leveraging Structural Prior and Constraints for Cryo-EM Heterogeneous Reconstruction 翻译

CryoSTAR: Leveraging Structural Prior and Constraints forCryo-EM Heterogeneous Reconstruction 翻译原文链接：https://www.biorxiv.org/content/10.1101/2023.10.31.564872v2.full.pdf

2024-11-21 14:10:13 908

原创 WHALE: TOWARDS GENERALIZABLE AND SCALABLE WORLD Models for Embodied Decision-making 翻译

WHALE: TOWARDS GENERALIZABLE AND SCALABLE WORLD Models for Embodied Decision-making 翻译

2024-11-21 13:28:47 970

原创 TOKENFORMER: RETHINKING TRANSFORMER SCAL- ING WITH TOKENIZED MODEL PARAMETERS 翻译

TOKENFORMER: RETHINKING TRANSFORMER SCAL- ING WITH TOKENIZED MODEL PARAMETERS 翻译

2024-11-21 13:17:58 1033

原创 How Far Is Video Generation from World Model: A Physical Law Perspective 翻译

How Far Is Video Generation from World Model: A Physical Law Perspective 翻译

2024-11-21 13:05:11 829

原创 The Surprising Effectiveness of Test-Time Training for Abstract Reasoning 翻译

The Surprising Effectiveness of Test-Time Training for Abstract Reasoning 翻译

2024-11-21 12:10:23 624

原创 Capturing forceful interaction with deformable objects using a deep learning- powered... 翻译

Capturing forceful interaction with deformable objects using a deep learning- powered stretchable tactile array 翻译

2024-11-21 11:50:53 811

原创 Scaling Laws for Precision 翻译

Scaling Laws for Precision 翻译

2024-11-21 11:28:16 1738

原创 That Chip Has Sailed: A Critique of Unfounded Skepticism Around AI for Chip Design 翻译

That Chip Has Sailed: A Critique of Unfounded Skepticism Around AI for Chip Design 翻译

2024-11-21 11:15:35 1990

原创 On Layer Normalization in the Transformer Architecture 翻译

The Transformer is widely used in natural language processing tasks. To train a Transformer however, one usually needs a carefully designed learning rate warm-up stage, which is shown to be crucial to the final performance but will slow down the optimizat-

2024-11-20 23:54:11 366

Doc2X的博客