A notebook, in public.
AI · Nanjing University

PERSONAL INDEX GitHub
Turing — On Computable Numbers, 1936Turing — On Computable Numbers, 1936 ↗ Lorenz attractorLorenz attractor ↗ Fourier series studyFourier series study ↗ Real, Symbolic, ImaginaryReal, Symbolic, Imaginary ↗ Mandelbrot setMandelbrot set ↗ Backpropagation and the matrix gradientBackpropagation and the matrix gradient ↗ Shannon — A Mathematical Theory of Communication, 1948Shannon — A Mathematical Theory of Communication, 1948 ↗ Multilayer perceptron — Dive into Deep LearningMultilayer perceptron — Dive into Deep Learning ↗ Gilbert Strang — Introduction to Linear AlgebraGilbert Strang — Introduction to Linear Algebra ↗ Walter Rudin — Principles of Mathematical AnalysisWalter Rudin — Principles of Mathematical Analysis ↗ 周志华 — 机器学习周志华 — 机器学习 ↗ Slavoj Žižek — The Sublime Object of IdeologySlavoj Žižek — The Sublime Object of Ideology ↗ Dive into Deep LearningDive into Deep Learning ↗ Structure and Interpretation of Computer ProgramsStructure and Interpretation of Computer Programs ↗ Alan TuringAlan Turing ↗ Geoffrey HintonGeoffrey Hinton ↗ Jacques LacanJacques Lacan ↗ King Crimson — In the Court of the Crimson King, 1969King Crimson — In the Court of the Crimson King, 1969 ↗ Neural Choreography — my animation projectNeural Choreography — my animation project ↗ GRU / LSTM — Neural ChoreographyGRU / LSTM — Neural Choreography ↗ Attention Is All You Need — Transformer architectureAttention Is All You Need — Transformer architecture ↗

A notebook, in public.

GitHub avatar of siddhartha

READ. BUILD. QUESTION. REPEAT.

Artificial intelligence undergraduate at Nanjing University.
AI, mathematics, computation, philosophy — and everything in between.

About me CV →References ↗

01Writing

All 24 essays →

RECENT ENTRIES

  1. 11 min read
    机器学习

    具身智能：榜单上的‘领先’到底是什么？

    从 RoboDojo 和 BEHAVIOR 的领先方案出发，讨论底座、经验获取与执行系统如何共同构成榜单成绩，以及这些经验能否帮助机器人处理偏差、迁移到新任务。

  2. 5 min read
    机器学习

    把神经网络连成一部动画

    用 Zanim 把《动手学深度学习》的 49 个片段连成一部动画，记下画面、转场和配乐的几轮返工。

  3. 3 min read
    哲学

    如果在夏夜，一个学生

    通过与自己的对话，反思功利、兴趣、理想与艺术，以及追求独特和超脱背后的自我表演。

  4. 1 min read
    其他

    个人向动画番剧最佳

    一张个人动画榜单：42 个心中的佳作、1 个入坑作，以及 6 个印象不太好的作品。

02Selected work

All projects →

Reinforcement learning

stickman-rl-lab

A 2D stickman learns to reach a target with PPO through an obstacle curriculum, with a desktop console for live training.

Full-course deterministic evaluation: 54/80 successes (67.5%). Natural upright walking is not claimed.

Python / PyTorch / Gymnasium / PyMunk

An articulated stickman and live PPO metrics in the desktop training console.
Training console · smoke run

Interactive learning

ai-anomaly-bureau

A browser investigation game about supervised learning, overfitting, hidden tests, and controlled experiments. Models train in the browser; test labels stay hidden until an audit.

React / TypeScript / Vite

Play the demo
A pixel-art cat is misclassified as bread in the opening scene of AI Anomaly Bureau.
Browser game · opening scene

Mathematical animation

neural-choreography

A continuous animation of neural-network mechanisms following Dive into Deep Learning: 49 segments, 11 min 21 s at 1080p60.

64 automated checks cover the underlying mathematics and rendering.

Python / Zanim / FFmpeg

Read the making-of →
Inputs, hidden ReLU neurons, and outputs in an animated multilayer perceptron.
Neural-network animation · still

GITHUB ACTIVITY SNAPSHOT

Activity snapshot.

@siddhartha-yz

1,161contributions

75active days

58 dayslongest streak

LAST 3 MONTHS

2 contributions on October 18, 2026

LessMore

Fixed snapshot from my public GitHub profile Updated