bluemsi/2nd
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: finalized wiki integrity maintenance (v3.0 standard) - pruned 1400+ stubs and fixed 11k+ ghost links

Browse Source
This commit is contained in:
Antigravity Agent
2026-05-02 09:18:34 +09:00
parent c84dcb8371
commit 6445fcc05b
13150 changed files with 55394 additions and 100862 deletions
Download Patch File Download Diff File Expand all files Collapse all files
10_Wiki/Topics_Blog/L2-Regularization.md
+2 -2
View File
@@ -6,7 +6,7 @@ tags: [auto-reinforced, l2-regularization, machine-learning, deep-learning, over
last_reinforced: 2026-04-20
---
# [[L2-Regularization]]
# [[L2-Regularization|L2-Regularization]]
## 📌 한 줄 통찰 (The Karpathy Summary)
> "신경망을 겸손하게: 모델이 특정 데이터에 너무 과하게 최적화(Overfitting)되어 괴물이 되지 않도록, 가중치값이 너무 커지면 벌금(Penalty)을 매겨 모델을 더 단순하고 부드럽게 만드는 수학적 억제제."
@@ -25,6 +25,6 @@ L2 정규화(L2-Regularization) 혹은 Ridge 정규화는 모델의 복잡도를
- **정책 변화(RL Update)**: 거대 모델 정책(Foundation-Models)에서는 파라미터가 너무 많아 정규화가 필수적이지만, 단순히 가중치를 줄이는 것을 넘어 '드롭아웃(Dropout)'이나 '데이터 증강' 등 다양한 정책과 혼합하여 사용됨.
## 🔗 지식 연결 (Graph)
- [[Gradient-Descent]], [[Optimization]], Deep Learning (DL), [[Efficiency]], Scaling-Laws
- [[Gradient-Descent|Gradient-Descent]], [[Optimization|Optimization]], Deep Learning (DL), [[Efficiency|Efficiency]], Scaling-Laws
- **Modern Tech/Tools**: Ridge regression, Weight decay in PyTorch/TensorFlow, AdamW optimizer.
---