© 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 24576
    9. 8-RES_FS24576-JB
    11. 12097
    Prev
    Next
    INDEX
    Explanations

    words related to geographic locations, specifically focusing on districts and states

    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
    430
    +0.12
    0.5%
    113
    +0.11
    0.5%
    288
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    113
    +0.12
    0.02
    537
    +0.11
    0.02
    31
    +0.11
    0.02
    Negative Logits
    cially
    -0.84
    cial
    -0.81
    ears
    -0.79
    cean
    -0.74
    aughs
    -0.74
    cius
    -0.74
    fty
    -0.73
    ndra
    -0.72
    offer
    -0.71
    cul
    -0.71
    POSITIVE LOGITS
     Attorney
    0.95
     District
    0.87
    s
    0.84
     Court
    0.81
     Marshal
    0.81
    rict
    0.77
     Clerk
    0.76
     Council
    0.72
     Circuit
    0.72
     Supervisor
    0.71
    Activations Density 0.015%

    No Known Activations