paper-replication

From-scratch PyTorch replications of classic and SOTA AI/ML papers: transformers, attention, GANs, diffusion, CLIP, LoRA, MoE and more, with reproducible training code.

my implementations of machine learning papers for learning and practice

Paper-to-code replication workspace with Jupyter notebooks and Python implementations of ML and deep learning research ideas.

📈 Independent replication and out-of-sample validation of Zarattini, Aziz & Barbon (2024) "Beat the Market" intraday momentum on SPY and ES futures, with ex-ante frozen anti-overfitting protocols.

Independent replication of the 5-minute Opening Range Breakout on QQQ (Zarattini & Aziz 2023, SSRN 4416622), stress-tested for execution costs. Break-even at ~2.2¢/share slippage; NQ confirmation filter significant per-trade (t=2.05) but 76% of its PnL is 2022 alone. Placebo control + bootstrap CIs.

Paper Replications

Evidence-first quantitative paper replications with claim ceilings and machine-checkable metadata

Gu, Kelly & Xiu (2020) reproduced on free data and a laptop GPU, with an Interactive Brokers paper-trading engine

Image-processing pipeline converting architectural drawings into binary line representations

PyTorch replication of Wang et al. (2018) "Deep Reinforcement Learning for Dynamic Multichannel Access." DQN vs. the analytical optimal policy on the round-robin switching case

Unofficial, From-scratch PyTorch replication of the Conformer paper (Gulati et al., 2020) — encoder, RNN-T decoder, training loop, and NeMo weight validation. Built block by block with documented maths.

Implementation of "Stock Options Trading Agents Guided by Option-Implied Return Distributions" (Xie & Liu, 2026), tested on 3 years of real option prices from $50k to $1M, with and without hedging. Option-implied features, 9 strategy families, resolvers, risk limits, backtester, Alpaca paper trading and a web UI for oversight.

Replication of SinKD: Sinkhorn Distance Minimization for Knowledge Distillation.

Learning Pytorch from learnpytorch.io, implementing on datasets.

Replication of "Neural Collaborative Filtering" (He et al., WWW 2017) in PyTorch — GMF, MLP & NeuMF on MovieLens 1M + Pinterest

This repository contains my PyTorch practice notebooks

Replication of the original "Learning Multi-level Features with Matryoshka SAEs" original paper

Jupyter notebook workspace for replicating ML paper ideas and experiments.

Hands-on replications of landmark AI papers — LeNet5, Word2Vec, Dropout & Federated Learning. From reading research to running it. Built with Python & PyTorch.

Replicating 'Reinforcement Learning from Language Feedback' (Klissarov et al., 2026) — Gemma 3 12B, GRPO, multi-turn teacher-student training on Omni-MATH