© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 17742
    Prev
    Next
    INDEX
    Explanations

    names or terms related to specific individuals

    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
    659
    +0.11
    0.5%
    496
    +0.10
    0.4%
    420
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.11
    0.02
    704
    +0.10
    0.01
    525
    +0.09
    0.01
    Negative Logits
    rontal
    -0.84
    unct
    -0.74
    ALLY
    -0.73
    ionic
    -0.71
    rious
    -0.67
    acts
    -0.67
    security
    -0.66
    ilateral
    -0.66
    ICAN
    -0.66
    ENTION
    -0.65
    POSITIVE LOGITS
    Anne
    0.93
     Chun
    0.83
    utenant
    0.82
    lee
    0.79
    wei
    0.78
     Harvey
    0.77
    vel
    0.76
     Stras
    0.76
    warm
    0.73
    pee
    0.72
    Activations Density 0.015%

    No Known Activations