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 24576
    5. 8-RES_FS24576-JB
    6. 8348
    Prev
    Next
    INDEX
    Explanations

    words related to specific actions or tools, like weapons or surveillance equipment

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    314
    +0.15
    0.7%
    409
    +0.11
    0.5%
    639
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.15
    0.04
    639
    +0.11
    0.03
    314
    +0.11
    0.01
    Negative Logits
     Tycoon
    -0.86
     Horus
    -0.73
     Misty
    -0.70
     Europa
    -0.70
     Daniels
    -0.69
     Schumer
    -0.68
     modification
    -0.68
     Dickinson
    -0.67
     Scotia
    -0.66
     Warn
    -0.66
    POSITIVE LOGITS
    paren
    1.20
    bodied
    1.07
    upon
    1.00
    up
    1.00
    looking
    0.99
    out
    0.99
    sounding
    0.99
    material
    0.97
    by
    0.95
    edged
    0.95
    Activations Density 0.052%

    No Known Activations