© 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. 9776
    Prev
    Next
    INDEX
    Explanations

    phrases related to political campaigns and finance

    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
    186
    +0.10
    0.5%
    595
    +0.10
    0.5%
    732
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.10
    0.07
    580
    +0.10
    0.06
    278
    +0.10
    0.04
    Negative Logits
    Buzz
    -0.69
    OTO
    -0.67
     PLEASE
    -0.65
     Crunch
    -0.63
    ico
    -0.62
     Dise
    -0.61
    wake
    -0.60
    pite
    -0.60
     reddit
    -0.60
    arat
    -0.58
    POSITIVE LOGITS
     indistinguishable
    0.91
    natureconservancy
    0.75
     predomin
    0.71
    emin
    0.71
     unthinkable
    0.69
     manifested
    0.63
    ided
    0.63
    ãĥĩãĤ£
    0.62
    arers
    0.62
    omorph
    0.61
    Activations Density 0.816%

    No Known Activations