© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 14898
    Prev
    Next
    INDEX
    Explanations

    words related to criminal activities and police investigations

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    710
    +0.14
    0.6%
    706
    +0.11
    0.5%
    459
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    710
    +0.14
    0.02
    670
    +0.11
    0.02
    706
    +0.10
    0.02
    Negative Logits
    ividual
    -0.68
    ertodd
    -0.68
    iaries
    -0.64
     Rabb
    -0.64
    ãħĭ
    -0.64
     faint
    -0.63
     reluct
    -0.62
    avorite
    -0.62
     Chinatown
    -0.61
     EntityItem
    -0.60
    POSITIVE LOGITS
    bral
    1.09
    xit
    1.08
    cht
    1.06
    cks
    1.04
    zza
    1.02
    versible
    0.98
    porter
    0.97
    ngth
    0.96
    gger
    0.95
    venge
    0.91
    Activations Density 0.031%

    No Known Activations