8000 Butanium (Clément Dumas) · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content
View Butanium's full-sized avatar
62 followers · 76 following

Achievements

Achievement: YOLOAchievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: Quickdraw

Achievements

Achievement: YOLOAchievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: Quickdraw

Highlights

  • Pro

Block or report Butanium

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.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, 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

Pinned Loading

  1. nnterp nnterp Public

    A small package implementing some useful wrapping around nnsight

    Python 13 3

  2. tiny-activation-dashboard tiny-activation-dashboard Public

    A tiny easily hackable implementation of a feature dashboard.

    Jupyter Notebook 10 2

  3. jkminder/dictionary_learning jkminder/dictionary_learning Public

    Forked from saprmarks/dictionary_learning

    Modified to support crosscoder training.

    Python 17 2

  4. jkminder/science-of-finetuning jkminder/science-of-finetuning Public

    Code for the "Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuning" paper.

    Jupyter Notebook 8 3

  5. llm-lang-agnostic llm-lang-agnostic Public

    minimal code to reproduce results from Separating Tongue from Thought: Activation Patching Reveals Language-Agnostic Concept Representations in Transformers

    Jupyter Notebook 7 1

  6. femtoGPT femtoGPT
    1 
    """
    2 
    A minimal pytorch implementation of a multi-head attention transformer inspired by nanoGPT
    3 
    """
    4 
    from dataclasses import dataclass
    5 
    import torch.nn as nn
0