© 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. 771
    Prev
    Next
    INDEX
    Explanations

    names of locations, particularly names of cities and football clubs

    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
    201
    +0.13
    0.6%
    481
    +0.12
    0.6%
    731
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    201
    +0.13
    0.06
    500
    +0.12
    0.05
    288
    +0.12
    0.06
    Negative Logits
    onym
    -0.93
    toggle
    -0.78
    atri
    -0.76
    aeda
    -0.74
    obin
    -0.73
    akable
    -0.73
    agy
    -0.73
    ographed
    -0.72
    Enlarge
    -0.71
    ajor
    -0.71
    POSITIVE LOGITS
    shire
    1.42
     Borough
    1.15
     County
    1.14
     City
    1.06
     Province
    0.99
     harbour
    0.98
     Airport
    0.98
     Region
    0.95
     Harbour
    0.94
     Labour
    0.93
    Activations Density 0.201%

    No Known Activations