BFS/A* Comparison

 

8 person (River Crossing 2)

20 persons1 (tree progression only) md, optimized share, artifact

search+scenario (hardcoded), (true animation of tree expansion2)

Comparing BFS 1&2

A* 8-20 side-by-side

Chain of Thought (CoT)

• Why we need chain of thought

• Chain of Thought vs. Cognition (concepts in slides with 3 IC design examples, enlarged)

Timing CoT Pipeline

Power CoT Pipeline

• Sonnet 4.6 vs. Extended Version, Performance Comparison
• Difference and Similarity between CoT & DFS, verbose

HW#1 Use AI Professionally

 

建議工具

使用 Claude Sonnet 4.6 推理模式(手動切換，免費用戶額定時間內只能使用三次)

使用 ChatGPT 5 推理模式(自動切換)

使用 Gemini 3.0 Pro 免費額度最高 1M tokens (永遠推理模式)

使用 Grok 4 推理模式(自動切換)

---

課堂練習 

Deadline: This Saturday at 23:59

Send all the share links to  me chang212@gmail.com by email with subject HW#1  [your id, your name]

How to publish a Claude artifact

How to share a ChatGPT link

How to share a Grok link

How to share Gemini Link

Content share 作業繳交格式

• share only link, pure text, markdown (md)
• no attachments accepted, no html, screen dump, or png
• non-compliant homework will be rejected and returned to you

任選一題

1. 

(a) Solve River Puzzles  (Pz 3, Pz 20) by Chain of Thought and Visualize/Animate the process

(b) Verify each step not violating constraints and correctness.

2. Comparing LLMs: Claude 4.6, Gemini 3.0 Pro, ChatGPT 5.0 (三選二) for the following task.

Animate in 3d (in three.js) Yagi-Uda Antenna, AI pipe inspector, or things you choose. （擇一）

The animation must be scientifically accurate. 

Hints

• Reasoning/Think/Extended Thinking Mode of the AI platform you use may be required for all the problems here
• Pz 3 solutions on various platforms
• Experiments on Pz 20 

River Crossing 20

 problem

animation, 2, 3d

Animation of thoughts, not just solution, 2

(包含問題分析、每個嘗試的死路、突破點、逐階段的約束驗證，以及最後的最優性證明和經驗總結。)

Analysis of strategies

Part II 

BFS Complexities 

BFS/A* search (share)

BFS/A* expansion

AI Platforms & Variations

 

• AI Animation比較 Claude vs. Gemini
• Opus vs. Sonnet

推薦 Opus 4.6 搭配思考模式（extended thinking）

我的使用體驗：

一次到位 vs. 反覆修正：這是 Opus 和 Sonnet 之間最核心的差異。Sonnet 在簡單任務上表現不錯，但面對複雜問題——像我們常做的 Three.js 地質動畫、LNA 教學投影片、A* 搜尋演算法視覺化這類高度整合的工作——Sonnet 容易在細節上出錯，需要來回修正好幾輪。Opus 因為推理深度更深，往往第一次就能把架構、邏輯、細節都處理到位。

Token 消耗的取捨：誠實地面對 Opus 單次消耗更多 token，這是因為思考模式下模型會在回應前進行更長的內部推理鏈。但從「總成本」角度來看，如果 Sonnet 要錯三次才修正到對，累積的 token 消耗和時間成本反而可能更高。這其實是一個「單次成本高但總效率高」vs.「單次便宜但需要多輪迭代」的經典取捨。

「比較一陣子才固定使用」：經過實際比較後的結論，更有參考價值。

AI Basics

 最好的程式語言 (Python? JavaScript?)

Prompt in English vs. Traditional English (中文版)

為什麼我的AI跟你不一樣: factors, basic skills，美學

supplemental

Claude 知識遭遇蒸餾事件

archive

提示使用英文的策略 (Word)

Open Remarks

JEPA World Model (scenarios)

 LNA IC Design

Portfolio Agents, roles

Trading

使用 Claude agent 回測 2025 

MSFT

AAPL

Google 

相較於 buy & hold
Agent 的表現
MSFT -1%
AAPL +22%
GOOGL +18%

Case Example: World Model for Artificial General Intelligence (AGI)

 (scenarios)

VLA vs. JEPA illustrated (Vision-Language-Action Model)

JEPA, an Introduction (slides in html)

Case Example

 

JEPA Visualization vs. DQN (Maze Doubled)

How the algorithm works and what the architecture looks like. 

JEPA learning

閱讀：JEPA issues within specific examples (html)

JEPA Double Pendulum  (share)

JEPA vs. Observation Space Reconstruction
Make it robust by adding noises (why, viz)

監控

maze

comparison of monitoring startegies

ps

AI 可以靠自己學習，探索，發現新事物嗎？  

Apple C1 Modem

A complete journey slides svg

 C1 Architecture

LNA schematic (TSMC N7 製程的完整設計流程，包含完整元件表、元件值推導公式、die 面積估算)

LNA Die

prompts

Compare GD, NM Simplex, and A*

NM Simplex (Nelder-Mead Simplex)

NM Simplex (animated 3d)

LNA Optimizer (SMITH Chart, Frequency Response), Merit of Figure (share)

Prompts, Cont. Prompts (share)

LNA Optimizer with Die Synced  (UI with synced sliders)

Prompts, Cont. Prompts (same as last)

Next day trading recommendation

 technical analysis

model + rule

slides