© 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 49152
    9. 8-RES_FS49152-JB
    11. 25720
    Prev
    Next
    INDEX
    Explanations

    the word "from" in various contexts related to events or occurrences

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    310
    +0.11
    0.5%
    373
    +0.10
    0.5%
    614
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    614
    +0.11
    0.02
    608
    +0.10
    0.03
    669
    +0.10
    0.02
    Negative Logits
    merce
    -0.89
    retty
    -0.73
    ICA
    -0.73
    ratulations
    -0.71
    video
    -0.70
    wcs
    -0.69
    spl
    -0.69
    inen
    -0.67
    endiary
    -0.67
    arella
    -0.67
    POSITIVE LOGITS
     unsuspecting
    1.02
     Palest
    0.85
     thence
    0.80
     afar
    0.80
     whence
    0.78
     underneath
    0.77
     beneath
    0.73
     somewhere
    0.72
     Rodney
    0.71
     abroad
    0.71
    Activations Density 0.051%

    No Known Activations