© 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 12288
    9. 8-RES_FS12288-JB
    11. 10724
    Prev
    Next
    INDEX
    Explanations

    mentions of specific geographic locations, particularly provinces and territories

    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
    551
    +0.11
    0.5%
    634
    +0.11
    0.5%
    219
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    219
    +0.11
    0.03
    612
    +0.11
    0.03
    288
    +0.10
    0.03
    Negative Logits
    paces
    -0.78
    ãĥīãĥ©
    -0.78
    geist
    -0.75
    chel
    -0.72
    ammad
    -0.71
    ihara
    -0.71
    ãĤ·ãĥ£
    -0.70
    Agent
    -0.70
    tle
    -0.70
    pter
    -0.69
    POSITIVE LOGITS
     Provincial
    0.95
     provincial
    0.91
     province
    0.86
     legislatures
    0.85
    incial
    0.81
     NDP
    0.80
     Scotia
    0.80
     premier
    0.79
     governments
    0.79
     parliament
    0.77
    Activations Density 0.082%

    No Known Activations