© 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. 1848
    Prev
    Next
    INDEX
    Explanations

    words related to hair care and processing, specifically focusing on instructions and precautions for bleaching hair

    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
    447
    +0.23
    1.1%
    314
    +0.12
    0.6%
    481
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    503
    +0.23
    0.05
    232
    +0.12
    0.06
    314
    +0.12
    0.01
    Negative Logits
     SHARES
    -0.75
     IMAGES
    -0.75
     Benz
    -0.74
     AMERICA
    -0.73
     Manor
    -0.70
     Deg
    -0.69
     Cruiser
    -0.69
     BI
    -0.68
     Butt
    -0.68
     iP
    -0.67
    POSITIVE LOGITS
    ission
    1.18
    posing
    1.16
    ading
    1.10
    estation
    1.09
    reating
    1.07
    ensing
    1.06
    amping
    1.06
    inding
    1.06
    rolling
    1.05
    iting
    1.04
    Activations Density 0.396%

    No Known Activations