Chrouos Home
Search
Ctrl
+
K
Chrouos Home
Search
Ctrl
+
K
Others
calendar
everyday
2026-02-25
2026-05-06
2026-05-08
Slides
Multiple Dialog (多輪對話) - Slides
WIDM
Masters Thesis
創建日誌－RAG with AutoPeerReview
Personal
Chrouo Obsidian Home Page
Personal Project
黃懷萱 Huai-Hsuan Huang
Reading
Algorithm
Aho-Corasick
KMP, Knuth-Morris-Pratt
Reverse Polish notation, RPN（逆波蘭表示法）
SequenceMatcher
二分搜尋法（Binary Search）
動態規劃（Dynamic Programming, DP）
Artificial Intelligence
Agents and Protocols
A2A（Agent-to-Agent Protocol）
Agent Markdown
Agent Skills
Agent Web Search (Outline)
Agent-FLAN, Designing Data and Methods of Effective Agent Tuning for Large Language Models
AI Agent (智慧代理)
AutoGPT
CrawAI
DuckDuckGo
Large Action Models, LAMs
LONGAGENT - Scaling Language Models to 128k Context through Multi-Agent Collaboration
Model Context Protocol, MCP
Multi-Agent System (MAS)
Reflection Agents
Reflexion - Language Agents with Verbal Reinforcement Learning
RepoAgent - An LLM-Powered Open-Source Framework for Repository-level Code Documentation Generation
SearXNG
Self-Reflection in LLM Agents Effects on Problem-Solving Performance
Tavily
Why Do Multi-Agent LLM Systems Fail
Architectures and Models
AI Image Generation (AI 生成圖片)
Auto-Encoder (自編碼器)
Cross-Attention (跨注意力機制)
Sliding Window Attention, SWA
語言模型的歷史（Language Model Evolution）
Classical ML Models
BAAI, Beijing Academy of Artificial Intelligence
Decision Tree (決策樹)
Gradient Boosted Tree(GBDT)
Naive Bayes Classifier (貝氏分類器)
Random Forest (隨機森林)
Support Vector Machine (SVM, 支持向量機)
Evaluation and Metrics
A Dataset of Peer Reviews (PeerRead) - Collection, Insights and NLP Applications
ACL Ready - RAG Based Assistant for the ACL Checklist
AgentReview🔍 - Exploring Academic Peer Review with LLM Agents
AUC
Automated Peer Reviewing in Paper SEA-Standardization, Evaluation, and Analysis
Automatic Analysis of Substantiation in Scientific Peer Reviews
BERTScore
BEVERS - A General, Simple, and Performant Framework for Automatic Fact Verification
BLEU
CHATEVAL - TOWARDS BETTER LLM-BASED EVALUATORS THROUGH MULTI-AGENT DEBATE
Confusion Matrix (混淆矩陣)
Evaluation Matrix
Is LLM a Reliable Reviewer A Comprehensive Evaluation of LLM on Automatic Paper Reviewing Tasks
LLMs Assist NLP Researchers - Critique Paper (Meta-)Reviewing
Mean Absolute Error, MAE (平均絕對誤差)
Mean Square Error, MSE (均方誤差)
New tools for understanding AI and learning outcomes
NLPEER - A Unified Resource for the Computational Study of Peer Review
REVIEWER2 - Optimizing Review Generation Through Prompt Generation
ROC, Receiver Operating Characteristic curve
Root Mean Square Error, RMSE (均方根誤差)
ROUGE（Recall-Oriented Understudy for Gisting Evaluation）
Feature Engineering
特徵工程 (feature-engineering)
Foundations
Classfication (分類)
Entropy
Regression (回歸)
Supervised Learning (監督式學習)
Unsupervised Learning (非監督式學習)
大型語言模型中的名詞解釋 ( LLM Glossary )
LLM-Models
ALBERT
BERT
DeBERTa
DistilBERT
ELECTRA
GPT-oss-120b stuck
mT5
OpenAI Prism
RoBERTa
MLOps
Ai2 Paper Finder
EasyChair
FLOPS, Floating Point Operations Per Second
GPU 雲端租用平台比較
MMA, Matrix Multiply and Accumulate
Nougat - Neural Optical Understanding for Academic Documents
Open Researcher and Contributor ID, ORCID
NLP
Auto Peer-Review
Context-1
Information Extraction
Multimodal Document-level Triple Extraction via Dynamic Graph Enhancement and Relation-Aware Reflection
Question Augmentation (問題增強)
SGP-TOD（Schema-Guided Prompting for Task-Oriented Dialog systems）
The Elephant in the Room - Analyzing the Presence of Big Tech in Natural Language Processing Research
Prompting and Context
Call Me When Necessary - LLMs can Efficiently and Faithfully Reason over Structured Environments
Chain-of-Thought（CoT）
Context Engineering
Context Rot (上下文腐爛)
Context Rot — LLM 越長越懂嗎
Decomposed Prompting - A Modular Approach for Solving Complex Tasks
In-Context Learning (上下文學習)
Is Chain-of-Thought Reasoning of LLMs a Mirage A Data Distribution Lens
LongContext Large Language Models
Prompt
TEACHING LARGE LANGUAGE MODELS TO SELFDEBUG
Teaching LLMs to reason like Bayesians
Tree of Thoughts (TOT)
切分策略（Chunking Strategy）
Regularization and Generalization
Critical Point (臨界點)
Error Surface (錯誤曲面)
L1 Regularization（Lasso）
L1 vs L2 正則化 — 實務差異
L2 Regularization（Ridge）
Local Minima（局部最小點）
Overfitting (過擬合)
Regularization (正則化)
Saddle Point（鞍點）
Underfitting (欠擬合)
Reinforcement Learning
DQN, Deep Q Network
MDP to Q-Learning to DQN (對照表)
MDP, Markov Decision Process
Q-Learning
RL, Reinforcement learning
Retrieval
Advanced RAG
Agentic RAG
ANN, Approximate Nearest Neighbor (近似最近鄰搜尋)
BM25
Cosine Similarity（Cosine Distance, 餘弦相似度）
CRAG, Corrective RAG
DAST. Context-Aware Compression in LLMs via Dynamic Allocation of Soft Tokens
DPR, Dense Vector Retrieval (稠密向量檢索)
Embedding Model
HNSW, Hierarchical Navigable Small World
LightRAG - Simple and Fast Retrieval-Augmented Generation
LLMs as a Reranker for RAG
LongLLMLingua - Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compression
Lp (Minkowski 距離族)
mmarco-mMiniLMv2-L12-H384-v1
Multilingual-E5-Large-Instruct
NLP 向量搜尋
RAG, Retrieval-Augmented Generation
Retrieving Support to Rank Answers in Open-DomainQuestion Answering
Self-RAG
Stronger Baselines for Retrieval-Augmented Generationwith Long-Context Language Models
Token
向量資料庫 (Vector Database)
Safety
Fairness Evaluation (公平性評估)
Guardrail AI
LLM Guardrail
Training and Optimization
Activation Function（激活函數）
Adaptive Learning Rate (適應性學習率)
Batch Normalization
Catastrophic Forgetting (災難性遺忘)
Cross-Entropy Loss（交叉熵損失）
Learning Rate (學習率)
Maximizing Likelihood (最大化似然)
Min-Max Scaling (最小－最大標準化)
Mode Collapse(模式坍塌)
Nested Learning, NL (嵌套學習)
QLORA - Efficient Finetuning of Quantized LLMs
Sigmoid
Softmax
Casual
同行評審 (PeerReview)
裁定更正 ( Judgment Correction )
Database
ETL, Extract, Transform and Load
Milvus Installation
DevOps
CQRS, Command Query Responsibility Segregation
DevOps
Microservices (微服務)
Monolithic Architecture (單體式架構)
Virtualization vs. Containerization
Notes
Crawler Journal
Paper
Paper Outline
Papers Outline
Python
Python - UV
Python 工程管理：最快理解速查
Security
2FA, Two-Factor Authentication (雙重驗證)
PKCE (Proof Key for Code Exchange)
Statistical
虛無假設 (Null Hypothesis, H0)
System
Conda & MiniConda
Docker
Github
Linux Operator
Tmux
Tools
LATEX 語法
Napkin 工具分享
Visual Studio Code
BERT
DistilBERT
Elasticsearch
FAISS（Facebook AI Similarity Search）
Guardrail AI
How Instacart Built a Search for Billions of Products
LLM Guardrail
Receptive Field (感受野)
RoBERTa
Simple Linear Regression
Enter your search text in the box above
Select a result to preview
Saddle Point（鞍點）
而Saddle Point則是左右兩側的Loss較高，前後的Loss較低，但由於Gradient為 0，所以模型不知道要往哪邊走
Powered by Forestry.md