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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 766
    Prev
    Next
    INDEX
    Explanations

    strings formatted as code snippets or software development 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_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
    156
    +0.12
    0.5%
    447
    +0.12
    0.5%
    502
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.03
    345
    +0.12
    0.04
    255
    +0.11
    0.04
    Negative Logits
    theless
    -0.74
     spoiler
    -0.73
     UNESCO
    -0.70
    "-
    -0.69
     Leopard
    -0.67
    ktop
    -0.65
     detrim
    -0.64
    00007
    -0.64
     afore
    -0.61
     Reich
    -0.61
    POSITIVE LOGITS
     =
    1.20
    _
    1.09
    Args
    0.93
    ={
    0.92
    Obj
    0.91
     :=
    0.89
    Value
    0.89
    func
    0.87
    (){
    0.87
    ){
    0.86
    Activations Density 0.147%

    No Known Activations