© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 6336
    Prev
    Next
    INDEX
    Explanations

    concepts related to features, benefits, and qualities of various subjects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    459
    +0.13
    0.6%
    218
    +0.10
    0.5%
    495
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    679
    +0.13
    0.12
    724
    +0.10
    0.12
    51
    +0.09
    0.10
    Negative Logits
     Tend
    -0.72
    ioxide
    -0.70
    ebook
    -0.68
     Annotations
    -0.68
     Alone
    -0.66
    itton
    -0.65
    otto
    -0.65
     Tsukuyomi
    -0.65
    enne
    -0.65
    ife
    -0.64
    POSITIVE LOGITS
     including
    1.08
     besides
    1.03
     imaginable
    0.93
     pertaining
    0.92
     relating
    0.91
     related
    0.90
    related
    0.85
     includ
    0.84
    including
    0.75
     associated
    0.73
    Activations Density 1.095%

    No Known Activations