zzqdeco ~/blog

#Deep Reading

共 33 篇 · 第 3 / 4 页
tags Deep Reading 33 Paper Reading 24 Code Reading 16 Computer Vision 12 Generative AI 5 Recommendation System 5 3D Generation 4 Agentic RL 3
还有 96 个标签 →
Deep Learning 3 Diffusion Models 3 Engineering 3 Image-to-3D 3 Knowledge Graph 3 Reinforcement Learning 3 Text-to-3D 3 3D Scene Generation 2 3D Scene Synthesis 2 Asynchronous RL 2 Computer Graphics 2 Graphics 2 NeRF 2 Optimization 2 RAG 2 Ranking 2 Systems 2 Two-Tower 2 Workflow 2 YouTube 2 3D Assets 1 3D Gaussian Splatting 1 3D Reconstruction 1 3D Scene Editing 1 3D Scene Modeling 1 3D Vision 1 Agent Memory 1 AI 1 AI Agents 1 Alibaba 1 ANN 1 Authorization 1 CAD Alignment 1 CAD Retrieval 1 Claude Code 1 Codex 1 Cold Start 1 Content Provenance 1 Context Engineering 1 Credit Assignment 1 CTR Prediction 1 Cypher 1 Database Internals 1 DeepSeek 1 Einops 1 Fourier Convolution 1 Generative Models 1 Generative Recommenders 1 Git 1 Graph Database 1 Graphiti 1 HSTU 1 Hybrid Search 1 Image Inpainting 1 Image Watermarking 1 Knowledge Augmented Generation 1 LLM Inference 1 LLM Post-Training 1 LLM Reliability 1 Long-Horizon RL 1 Loss Landscape 1 Machine Learning 1 Meta 1 Multi-Task Learning 1 Multimodal LLM 1 Multitask Learning 1 Neo4j 1 Notes 1 OpenFGA 1 OpenSPG 1 Pose Estimation 1 Python 1 Qwen 1 ReBAC 1 Rectified Flow 1 Referring Segmentation 1 Search 1 Search Engine 1 Security 1 Semi-Autoregressive Generation 1 Sequential Recommendation 1 Single-view Reconstruction 1 Speculative Decoding 1 Survey 1 System Design 1 Tensor Programming 1 Theory 1 Training Systems 1 Tutorial 1 User Interest Modeling 1 Vector Search 1 Vespa 1 Visual Grounding 1 Works 1 Zanzibar 1 Zep 1

2026

2026.06

  1. Blended-NeRF 论文精读：现有 NeRF 场景中的零样本对象生成、ROI 编辑与体渲染融合

    精读 Blended-NeRF 如何通过 3D ROI、CLIP 引导、双 NeRF 场融合、距离平滑和官方代码实现现有神经辐射场中的对象插入、替换与材质编辑

    Deep ReadingPaper ReadingNeRF +5

  2. CompoNeRF 论文精读：文本引导多物体组合 NeRF、可编辑 3D 布局与场景重组

    精读 CompoNeRF 如何通过局部/全局文本引导、密度组合、可编辑 3D box layout 与 NeRF 分解重组生成多物体 3D 场景

    Deep ReadingPaper ReadingText-to-3D +4

  3. DSpark 论文精读：置信度调度推测解码、半自回归生成与大模型推理加速

    精读 DeepSeek DSpark 如何通过半自回归 draft model、confidence-scheduled verification 和 DeepSpec 训练评估代码改进大模型 speculative decoding 的速度、接受率与吞吐边界

    Deep ReadingPaper ReadingLLM Inference +4

  4. Generative Recommenders 论文精读：HSTU、M-FALCON 与万亿参数推荐序列建模

    精读 Meta ICML 2024 Generative Recommenders 如何将推荐系统重构为序列生成问题，并通过 HSTU、Stochastic Length、M-FALCON 和 scaling law 支撑工业级部署

    Deep ReadingPaper ReadingRecommendation System +4

  5. Clippy 论文精读：多任务推荐排序模型的训练稳定性、损失发散与自适应更新裁剪

    精读 KDD 2023 Clippy 如何分析 YouTube 多任务排序模型训练不稳定问题，并用 Adagrad 更新裁剪提升大规模推荐模型稳定性

    Deep ReadingPaper ReadingRecommendation System +4

  6. YouTube DNN 推荐系统论文精读：双阶段召回排序、候选生成与观看时长优化

    精读 RecSys 2016 YouTube DNN 推荐系统如何用深度候选生成、排序模型、样本构造、特征工程和观看时长目标支撑大规模推荐

    Deep ReadingPaper ReadingRecommendation System +4

  7. 简单推荐算法实施指南：从冷启动弱监督到真实反馈闭环

    系统梳理推荐系统在缺少样本时如何通过内容先验、规则蒸馏、受控探索、曝光日志和双塔召回逐步建立稳定训练闭环

    Deep ReadingRecommendation SystemCold Start +4

  8. Imaginarium 论文精读：视觉引导、高质量资产库、代码实现与 3D 场景布局生成

    从论文与官方代码双线精读 Imaginarium 如何通过 Flux 图像引导、视觉解析、资产检索、姿态估计、场景图约束和 Blender 物理优化生成高质量 3D 场景布局

    Deep ReadingPaper Reading3D Scene Synthesis +4

  9. CasLayout 论文精读：级联 3D 布局扩散、稀疏关系建模与室内场景生成

    精读 CasLayout 如何通过四阶段条件扩散、建筑约束、稀疏关系图与 Relation VAE 实现可控 3D 室内布局生成

    Deep ReadingPaper Reading3D Scene Synthesis +3

  10. 长期收益强化学习算法研究进展：从时间信用分配到 Agentic RL

    系统梳理长时域、稀疏延迟奖励强化学习中的信用分配、时间抽象、离线到在线、世界模型与 LLM Agent RL 进展

    Deep ReadingReinforcement LearningLong-Horizon RL +3