© 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. 10-RES_POST_32K-OAI
    11. 29040
    Prev
    Next
    INDEX
    Explanations

    numerical expressions and computational syntax

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    160
    +0.17
    0.8%
    107
    +0.13
    0.6%
    378
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    731
    +0.17
    0.09
    254
    +0.13
    0.11
    107
    +0.13
    0.13
    Negative Logits
     Parables
    -0.80
     eleph
    -0.78
     seiz
    -0.74
    glomer
    -0.73
     TAMADRA
    -0.71
     princ
    -0.69
     pione
    -0.69
     Wast
    -0.68
     Manit
    -0.68
     Marketable
    -0.67
    POSITIVE LOGITS
    [
    1.09
    ERROR
    1.07
    <
    1.06
    {
    1.05
    Reference
    1.01
    None
    1.00
    []
    1.00
    âĪĴ
    1.00
    ["
    0.99
    Default
    0.99
    Activations Density 1.057%

    No Known Activations