Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 768
    5. 8-RES_FS768-JB
    6. 326
    Prev
    Next
    INDEX
    Explanations

    numerical and measurement-related terms and concepts

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    481
    +0.11
    0.5%
    255
    +0.10
    0.5%
    447
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.23
    575
    +0.10
    0.18
    455
    +0.10
    0.17
    Negative Logits
    arus
    -0.66
    lé
    -0.65
    verbs
    -0.65
    pload
    -0.61
    bows
    -0.60
    urated
    -0.60
    hran
    -0.60
    wolves
    -0.59
    ipeg
    -0.59
    actionDate
    -0.59
    POSITIVE LOGITS
     thereof
    1.15
     of
    1.02
     guiActiveUnfocused
    0.73
     onwards
    0.72
    icals
    0.71
    ":[
    0.71
     comprising
    0.69
    .(
    0.67
    naires
    0.66
     (~
    0.65
    Activations Density 5.339%

    No Known Activations