© 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 49152
    9. 8-RES_FS49152-JB
    11. 25987
    Prev
    Next
    INDEX
    Explanations

    words related to specific actions and processes, particularly in the context of crime and investigation

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    309
    +0.14
    0.6%
    373
    +0.12
    0.5%
    616
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    309
    +0.14
    0.02
    284
    +0.12
    0.01
    373
    +0.09
    -0.01
    Negative Logits
     Jagu
    -0.74
     Gret
    -0.72
     Spock
    -0.70
     Haas
    -0.70
     Kardash
    -0.69
    glers
    -0.66
     Kru
    -0.64
     NK
    -0.64
     Kardashian
    -0.63
     Rutgers
    -0.62
    POSITIVE LOGITS
    ophen
    0.94
    ritic
    0.81
    ONY
    0.81
    ornia
    0.80
    itect
    0.80
    irl
    0.79
    athy
    0.78
    gypt
    0.77
    emy
    0.77
    emies
    0.76
    Activations Density 0.024%

    No Known Activations