1. Installing Emacs on Win7/Win8 for dummies
經常要幫人安裝emacs，寫給自己的備忘錄；基本上適用於 windows 系列的作業系統
2. Elpa 的其它程式庫安裝，請參考 http://www.emacswiki.org/emacs/ELPA
skip to main |
skip to sidebar
我們選擇在這十年內登陸月球並完成其他的事，不是因為他們很簡單，而是因為他們很艱難。~~約翰·甘迺迪
搜尋此網誌
熱門文章
-
本次訪談邀請到深耕人工智慧與軟體工程多年的Andre Karpathy，討論AI代理（agents）技術的革命性進展、軟體工程與研究自動化的未來願景，以及AI在家居自動化、教育和機器人領域的應用趨勢。 AI代理顛覆軟體開發 Karpathy 表示，從2023年12月開始，他...
-
這段影片介紹了一位創作者如何利用Andre Karpathy（前特斯拉AI主管、OpenAI創始成員）發布的自動研究（Auto Research）開源專案，打造一套能自我提升的內容產製機器人。該系統每天產出5支影片，發布後根據Instagram及Facebook的真實觀看數據，自...
-
這段訪談深入探討了宇宙可能是模擬運行的假設，並涉及多領域專家的觀點，包括遊戲設計、哲學、物理學、心理學與生命科學。 首先，卡內基梅隆大學的遊戲設計專家Jesse Shell指出，雖然當代遊戲無法完美模擬三維世界與複雜感官體驗，但透過簡化「主動玩家」與「背景角色」的運算方式，可...
-
本集《Y Files》深入探討了一種被稱為「時間風暴」的超自然現象，這類事件在全球範圍內已被記錄數百年，許多文化對其有不同的稱呼。時間風暴被描述為帶有光芒的彩色迷霧，能引起時間感知扭曲、失蹤甚至物理變化的神秘現象。 故事從1977年智利山區的軍士瓦爾德茲（Armando Va...
-
在大型語言模型（Large Language Models, LLM）持續推進自然語言處理能力的今天，如何有效且深入理解這些模型在微調（Finetuning）過程中內部行為的演變，成為一個極具挑戰且重要的議題。Ren 與 Sutherland 於 ICLR 2025 發表的論文...
-
在機器學習與統計建模領域中，對離散分布的有效抽樣（sampling）一直是一項挑戰。尤其是在高維度與複雜結構的概率模型中，如何利用梯度資訊改善抽樣效率仍未有明確解決方案。Grathwohl 等人在 2021 年 ICML 發表的論文《Oops I Took A Gradient...
-
在深度學習蓬勃發展的當下，模型架構愈來愈複雜，而如何使模型在取得高表現的同時，兼顧運算效率與彈性，成為研究重點。由 Chen 等人於 NeurIPS 2018 發表的 Neural Ordinary Differential Equations (Neural ODEs) ，提...
-
在深度學習領域中，多層感知器（Multi-Layer Perceptrons，簡稱 MLP）長期以來都是神經網路的基石模型，其架構包括固定形式的節點激活函數與線性權重參數。然而，ICLR 2024 年榮獲 Outstanding Paper 獎的論文《KAN: Kolmogor...
-
主成分分析（PCA）是機器學習與資料科學中極為重要的維度約減技術，廣泛應用於資料前處理、特徵萃取、降噪及資料視覺化等領域。傳統的 PCA 通常透過特徵分解（eigen decomposition）或奇異值分解（SVD）求解最大方差方向。然而，隨著資料規模的爆炸性成長以及分散式計...
-
隨著機器學習技術日益滲透社會各個層面，公平性（fairness）問題成為研究熱點，尤其是在分類決策系統中如何避免對少數族群或弱勢群體產生歧視。然而，大多數公平性研究及其評估標準主要集中在靜態的決策階段，忽略了這些決策如何隨時間影響整體族群的福祉與長期發展。ICML 2018 年...
網誌存檔
-
▼
2026
(413)
-
▼
4月
(213)
- Learning Dynamics of LLM Finetuning — 解析大型語言模型微調的學習動態
- Safety Alignment Should Be Made More Than Just a F...
- KAN: Kolmogorov-Arnold Networks — 以科爾莫哥洛夫-阿諾爾德表示定理...
- Mamba: Linear-Time Sequence Modeling with Selectiv...
- Learning Interactive Real-World Simulators
- Generalization in Diffusion Models Arises from Geo...
- Rethinking the Expressive Power of GNNs via Graph ...
- Universal Few-shot Learning of Dense Prediction Ta...
- DreamFusion: 利用 2D 擴散模型實現文字到 3D 的生成革命
- 基於流網路的生成模型：非迭代多樣候選生成之GFlowNet方法深入解析
- Hyperparameter Tuning with Renyi Differential Priv...
- Analytic-DPM: an Analytic Estimate of the Optimal ...
- Learning Mesh-Based Simulation with Graph Networks...
- EigenGame: PCA as a Nash Equilibrium — 一種以賽局理論視角重新...
- An Image is Worth 16x16 Words: Transformers for Im...
- Ordered Neurons: Integrating Tree Structures into ...
- The Lottery Ticket Hypothesis: Finding Sparse, Tra...
- Continuous Adaptation via Meta-Learning in Nonstat...
- Spherical CNNs：在球面上的卷積神經網路革命
- On the Convergence of Adam and Beyond
- Discrete Diffusion Modeling by Estimating the Rati...
- Measure Dataset Diversity, Don't Just Claim It：從測量...
- Position: Considerations for Differentially Privat...
- A Watermark for Large Language Models — ICML 2023 ...
- Learning-Rate-Free Learning by D-Adaptation
- G-Mixup: Graph Data Augmentation for Graph Classif...
- Understanding Self-Supervised Learning Dynamics wi...
- Oops I Took A Gradient: Scalable Sampling for Disc...
- Unbiased Gradient Estimation in Unrolled Computati...
- Generative Pretraining From Pixels (iGPT) 深度簡介
- On Learning Sets of Symmetric Elements 深度簡介
- Rates of Convergence for Sparse Variational Gaussi...
- Challenging Common Assumptions in the Unsupervised...
- Delayed Impact of Fair Machine Learning 深度簡介
- Obfuscated Gradients Give a False Sense of Securit...
- 以自身不完美版本引導擴散模型：提升影像品質與多樣性的創新方法解析
- Visual Autoregressive Modeling: Scalable Image Gen...
- DecodingTrust：全面評估 GPT 模型的可信度
- Direct Preference Optimization: Your Language Mode...
- Scaling Data-Constrained Language Models
- Privacy Auditing with One (1) Training Run - 深度解析
- Are Emergent Abilities of Large Language Models a ...
- Riemannian Score-Based Generative Modelling 深度簡介
- Gradient Descent: The Ultimate Optimizer — NeurIPS...
- Is Out-of-Distribution Detection Learnable? —— Neu...
- Elucidating the Design Space of Diffusion-Based Ge...
- On the Expressivity of Markov Reward
- A Universal Law of Robustness via Isoperimetry 深度解析
- Improved Guarantees and a Multiple-Descent Curve f...
- No-Regret Learning Dynamics for Extensive-Form Cor...
- Language Models are Few-Shot Learners (GPT-3) — 深度簡介
- Uniform convergence may be unable to explain gener...
- Distribution-Independent PAC Learning of Halfspace...
- Nearly Tight Sample Complexity Bounds for Learning...
- Optimal Algorithms for Non-Smooth Distributed Opti...
- Non-delusional Q-learning and Value-iteration
- Neural Ordinary Differential Equations
- A Linear-Time Kernel Goodness-of-Fit Test 深度介紹
- Safe and Nested Subgame Solving for Imperfect-Info...
- Superposition Yields Robust Neural Scaling 深度解析
- Optimal Mistake Bounds for Transductive Online Lea...
- Does Reinforcement Learning Really Incentivize Rea...
- Why Diffusion Models Don't Memorize: The Role of I...
- 1000 Layer Networks for Self-Supervised RL: Scalin...
- Gated Attention for Large Language Models: Non-lin...
- Artificial Hivemind: The Open-Ended Homogeneity of...
- The Value of Prediction in Identifying the Worst-Off
- Score Matching with Missing Data 深度解析
- Conformal Prediction as Bayesian Quadrature
- Roll the Dice & Look Before You Leap: Going Beyond...
- Train for the Worst, Plan for the Best: Understand...
- CollabLLM: From Passive Responders to Active Colla...
- Data Shapley in One Training Run：高效精準的資料貢獻度評估新範式
- Learning Dynamics of LLM Finetuning 深度解析
- Safety Alignment Should Be Made More Than Just a F...
- KAN: Kolmogorov-Arnold Networks — 打破傳統神經網路結構的全新架構探索
- Mamba: Linear-Time Sequence Modeling with Selectiv...
- Learning Interactive Real-World Simulators
- Generalization in Diffusion Models Arises from Geo...
- Rethinking the Expressive Power of GNNs via Graph ...
- Universal Few-shot Learning of Dense Prediction Ta...
- DreamFusion: 使用 2D 擴散模型實現文字轉 3D 的突破性方法
- Flow Network based Generative Models for Non-Itera...
- Hyperparameter Tuning with Renyi Differential Priv...
- Analytic-DPM: an Analytic Estimate of the Optimal ...
- Learning Mesh-Based Simulation with Graph Networks...
- EigenGame: PCA as a Nash Equilibrium 深度解析
- An Image is Worth 16x16 Words: Transformers for Im...
- Ordered Neurons: Integrating Tree Structures into ...
- The Lottery Ticket Hypothesis: Finding Sparse, Tra...
- Continuous Adaptation via Meta-Learning in Nonstat...
- Spherical CNNs：在球面上的卷積神經網路革新
- On the Convergence of Adam and Beyond
- Discrete Diffusion Modeling by Estimating the Rati...
- Measure Dataset Diversity, Don't Just Claim It
- Position: Considerations for Differentially Privat...
- A Watermark for Large Language Models 深度簡介
- Learning-Rate-Free Learning by D-Adaptation - ICML...
- G-Mixup: Graph Data Augmentation for Graph Classif...
- Understanding Self-Supervised Learning Dynamics wi...
-
▼
4月
(213)
沒有留言:
張貼留言