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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 1504
    Prev
    Next
    INDEX
    Explanations

    news articles and blog posts mentioning specific names and events related to politics and current affairs, written by specific authors

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    532
    +0.11
    0.5%
    447
    +0.10
    0.5%
    426
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    609
    +0.11
    0.06
    76
    +0.10
    0.06
    573
    +0.10
    0.04
    Negative Logits
    cale
    -0.75
    ignt
    -0.74
    glomer
    -0.71
    perty
    -0.70
    anmar
    -0.67
    FTWARE
    -0.64
     challeng
    -0.63
    sbm
    -0.62
    undai
    -0.62
    ãģĦ
    -0.62
    POSITIVE LOGITS
     Details
    0.96
    :
    0.90
     Notes
    0.83
     Description
    0.82
     Highlights
    0.79
     #
    0.79
     Thumbnails
    0.77
     courtesy
    0.75
     Explan
    0.74
     wise
    0.74
    Activations Density 0.457%

    No Known Activations