牛哥精选 · 半年

1

📝 深度技术 arXiv 计算机视觉 2026-08-19

Hierarchical Data Selection via Manifold Coverage and Sparse Feature Coverage in LLM Post-training

流形覆盖结合稀疏特征覆盖，为LLM后训练数据筛选提供分层新思路。

arXiv:2608.16927v1 Announce Type: cross Abstract: As supervised fine-tuning data continues to scale, selecting high-value subsets from large candidate…

数据选择大模型后训练流形覆盖稀疏特征 llm

2

📝 深度技术 arXiv AI 2026-08-04

Divisive Normalization Shapes Low-Rank Slow Manifolds for Continuous Working Memory

揭秘大脑连续工作记忆的数学机制，看归一化如何塑造低秩慢流形。

arXiv:2608.01947v1 Announce Type: cross Abstract: The ability to robustly maintain and update continuous variables is a hallmark of working memory. Wh…

工作记忆神经科学计算模型归一化低秩结构

3

📝 深度技术 arXiv AI 2026-07-22

Riemannian Deep Learning:Modules, Networks, and Geometries

突破欧式空间局限，探索黎曼流形上的深度学习框架，为几何学习提供新范式。

arXiv:2607.19305v1 Announce Type: cross Abstract: Deep neural networks on manifold-valued representations have attracted growing interest, but many ba…

riemannian 流形几何神经网络结构几何机器学习

4

📝 深度技术 arXiv 机器学习 2026-07-07

MANCE: Manifold Aware Concept Erasure

利用流形结构精准擦除特定概念，为AI安全与可解释性开辟新路径。

arXiv:2607.03973v1 Announce Type: new Abstract: Concept erasure aims to remove a target concept from a representation while preserving the other infor…

mance 流形意识概念擦除流形学习可解释ai

5

🤖 AI 工具 arXiv 机器学习 2026-07-02

Diffeomorphic Optimization

用微分同胚映射优化高维流形，破解非凸损失下的生成模型训练难题，让梯度下降目标更清晰。

arXiv:2607.00947v1 Announce Type: new Abstract: Generative models learn data distributions that reside on a low-dimensional manifold within a higher-d…

微分同胚流形优化生成模型非凸损失梯度下降

6

📝 深度技术 arXiv 计算机视觉 2026-06-30

Towards Modality-Agnostic Medical Image Anomaly Detection: A Training-Free Manifold Refinement Approach

不用训练即可适配多种医学影像模态，流形细化让异常检测更高效精准。

arXiv:2604.19191v2 Announce Type: replace Abstract: Deploying AI-based anomaly detection across diverse clinical imaging settings remains challenging …

医学图像异常检测流形学习多模态无训练

7

📝 深度技术 arXiv 计算机视觉 2026-06-18

On-Manifold Variational Learning with Heat-Kernel Priors

将热核先验引入流形变分框架，为贝叶斯深度学提供新的几何先验设计思路。

arXiv:2606.18658v1 Announce Type: new Abstract: Learning unsupervised representations of medical imaging cohorts can reveal clinically meaningful prot…

流形学习变分推理热核先验贝叶斯深度学习几何深度学习

8

🤖 AI·大模型 arXiv AI 2026-06-10

Mitigating Manifold Departure: Uncertainty-Aware Subspace Rectification for Trustworthy MLLM Decoding

ICML 2026新研究用不确定性感知子空间纠正，让多模态大模型解码更可信，有效缓解流形偏离问题。

arXiv:2606.09859v1 Announce Type: cross Abstract: MLLMs frequently hallucinate objects inconsistent with visual inputs. This issue is typically attrib…

多模态大模型不确定性量化流形学习可信解码子空间修正

9

📝 深度技术 arXiv 计算机视觉 2026-06-02

Visualizing definitional divergence in high-dimensional data by manifold alignment: Application to 3D right ventricular strain computations

流形对齐高维数据可视化，解决定义分歧，提升3D右心室应变计算精度

arXiv:2501.12178v2 Announce Type: replace Abstract: Medical imaging studies often rely on a single sample per subject, assuming it is representative o…

高维数据可视化流形对齐定义分歧右心室应变医学图像分析

10

📝 深度技术 arXiv 机器学习 2026-05-21

LLM Pretraining Shapes a Generalizable Manifold: Insights into Cross-Modal Transfer to Time Series

LLM预训练的隐藏能力：学习到的数据流形可跨模态迁移至时间序列任务，揭示通用表征机制。

arXiv:2605.20449v1 Announce Type: new Abstract: Can language-pretrained transformers become effective time-series forecasters, and why? In this paper,…

llm 预训练跨模态迁移时间序列流形学习

11

📝 深度技术 arXiv 机器学习 2026-05-20

Transformers for Learning on Noisy and Task-Level Manifolds: Approximation and Generalization Insights

从理论层面揭示Transformer在噪声与任务级流形上的学习能力，近似与泛化分析带来新洞察

arXiv:2505.03205v3 Announce Type: replace Abstract: Transformers serve as the foundational architecture for large language and video generation models…

transforme 噪声流形任务级流形泛化理论近似理论

12

📝 深度技术 arXiv 机器学习 2026-05-20

Riemannian Networks over Full-Rank Correlation Matrices

被ICML 2026接收，提出在全秩相关矩阵上构建黎曼网络的新几何深度学习方法。

arXiv:2605.19073v1 Announce Type: new Abstract: Representations on the Symmetric Positive Definite (SPD) manifold have garnered significant attention …

黎曼网络相关矩阵 icml 2026 深度几何学习黎曼几何

13

📝 深度技术 arXiv 机器学习 2026-05-20

Manifold Random Features

探索流形上的随机特征方法，为核方法在复杂数据空间中的应用提供新思路。

arXiv:2602.03797v3 Announce Type: replace Abstract: We present a new paradigm for creating random features to approximate bi-variate functions (in par…

流形学习随机特征核方法 arxiv论文理论创新

14

📝 深度技术 arXiv 机器学习 2026-05-20

Spherical Harmonic Optimal Transport: Application to Climate Models Comparisons

利用球谐函数加速最优传输计算，降低流形上成本，精准比较气候模型差异。

arXiv:2605.18389v1 Announce Type: new Abstract: Optimal transport provides a powerful framework for comparing measures while respecting the geometry o…

球谐函数最优传输气候模型流形计算成本

15

📝 深度技术 arXiv 机器学习 2026-05-19

The Geometric Structure of Models Learning Sparse Data

探究机器学习中流形假设的局限性，揭示模型学习稀疏数据时的几何结构新发现。

arXiv:2605.08464v2 Announce Type: replace Abstract: The manifold hypothesis (MH) is often used to explain how machine learning can overcome the curse …

稀疏数据流形假设机器学习几何结构高维数据

16

📝 深度技术 arXiv 机器学习 2026-05-19

Mind Dreamer: Untethering Imagination via Active Latent Intervention on Latent Manifolds

MBRL新突破：主动潜在干预打破历史束缚，让想象力从潜在流形自由生成

arXiv:2605.16030v1 Announce Type: new Abstract: Model-Based Reinforcement Learning (MBRL) leverages latent imagination for sample efficiency, yet rema…

强化学习模型驱动潜在想象主动干预潜在流形

🐂 牛哥精选

Hierarchical Data Selection via Manifold Coverage and Sparse Feature Coverage in LLM Post-training

Divisive Normalization Shapes Low-Rank Slow Manifolds for Continuous Working Memory

Riemannian Deep Learning:Modules, Networks, and Geometries

MANCE: Manifold Aware Concept Erasure

Diffeomorphic Optimization

Towards Modality-Agnostic Medical Image Anomaly Detection: A Training-Free Manifold Refinement Approach

On-Manifold Variational Learning with Heat-Kernel Priors

Mitigating Manifold Departure: Uncertainty-Aware Subspace Rectification for Trustworthy MLLM Decoding

Visualizing definitional divergence in high-dimensional data by manifold alignment: Application to 3D right ventricular strain computations

LLM Pretraining Shapes a Generalizable Manifold: Insights into Cross-Modal Transfer to Time Series

Transformers for Learning on Noisy and Task-Level Manifolds: Approximation and Generalization Insights

Riemannian Networks over Full-Rank Correlation Matrices

Manifold Random Features

Spherical Harmonic Optimal Transport: Application to Climate Models Comparisons

The Geometric Structure of Models Learning Sparse Data

Mind Dreamer: Untethering Imagination via Active Latent Intervention on Latent Manifolds

📅 日期