Skip to content
View ydnyshhh's full-sized avatar

Highlights

  • Pro

Block or report ydnyshhh

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ydnyshhh/README.md

Project banner

Hi, I'm Yadnyesh — an aspiring Machine Learning Researcher with a deep passion for understanding and advancing AI. My research interests include Diffusion Models, Reinforcement Learning (RL), Mechanistic Interpretability , Scientific Machine Learning, and Foundational Large Language Models (LLMs). I am particularly interested in building interpretable, efficient, and scalable AI systems that contribute to scientific discovery and create real‑world impact.

Pinned Loading

  1. smol-moe smol-moe Public

    smol-moe is a minimal and modular implementation of a Sparse Mixture of Experts (MoE) transformer in PyTorch.

    Python 1

  2. ydnyshhh.github.io ydnyshhh.github.io Public

    HTML

  3. fine-pipe fine-pipe Public

    fine-pipe converts pasted web links into a structured, pretraining-ready dataset. It performs robust fetching, article-grade text extraction, language and quality filtering, MinHash deduplication, …

    Python

  4. align-drift align-drift Public

    Align-Drift measures how post-training (SFT, DPO, GRPO) shifts an LLM’s behavior and internal representations. It uses CKA/spectral-rank probes and mechanistic patching/ablation to localize where a…

    Jupyter Notebook