© 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. 14897
    Prev
    Next
    INDEX
    Explanations

    details of criminal charges and arrests

    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
    604
    +0.09
    0.4%
    613
    +0.09
    0.4%
    295
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.09
    0.05
    295
    +0.09
    0.03
    604
    +0.09
    0.03
    Negative Logits
    ankind
    -0.79
     nostalg
    -0.76
    iuses
    -0.72
    warts
    -0.72
    wolves
    -0.71
    sson
    -0.68
     envy
    -0.67
     iceberg
    -0.64
    sels
    -0.64
    NetMessage
    -0.63
    POSITIVE LOGITS
     Detention
    1.11
     probation
    1.06
     pending
    1.03
     arra
    1.00
     Correctional
    0.99
     misdemeanor
    0.96
     felony
    0.91
     custody
    0.90
     Corrections
    0.90
     criminally
    0.89
    Activations Density 0.214%

    No Known Activations