© 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. 15569
    Prev
    Next
    INDEX
    Explanations

    specific locations or dates

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    110
    +0.11
    0.5%
    151
    +0.10
    0.5%
    693
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    110
    +0.11
    0.06
    693
    +0.10
    0.06
    288
    +0.10
    0.06
    Negative Logits
    soType
    -0.82
    thia
    -0.75
    enth
    -0.73
     oneself
    -0.72
    vier
    -0.71
    urtles
    -0.70
    uador
    -0.70
    hovah
    -0.69
    aples
    -0.68
    beit
    -0.68
    POSITIVE LOGITS
     counterparts
    1.53
     counterpart
    1.50
     namesake
    1.29
     brethren
    1.13
     colleague
    1.12
     portfolio
    1.06
     persona
    1.04
     heritage
    1.02
     predecessors
    1.02
     cousin
    1.01
    Activations Density 0.390%

    No Known Activations