© 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. 15026
    Prev
    Next
    INDEX
    Explanations

    phrases related to political and personal statements

    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
    674
    +0.10
    0.5%
    284
    +0.10
    0.5%
    209
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    398
    +0.10
    0.05
    482
    +0.10
    0.04
    284
    +0.10
    0.05
    Negative Logits
    ONSORED
    -0.79
    >[
    -0.76
    iven
    -0.75
    ":""},{"
    -0.74
    ":["
    -0.73
    relevant
    -0.71
    iors
    -0.70
    SPA
    -0.70
    :[
    -0.69
    employed
    -0.69
    POSITIVE LOGITS
     gown
    0.98
     mortar
    0.96
     cheese
    0.85
     brim
    0.84
     shovel
    0.79
     feather
    0.78
     feathers
    0.77
     stone
    0.77
     gravy
    0.76
     Barrel
    0.76
    Activations Density 0.165%

    No Known Activations