© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 9944
    Prev
    Next
    INDEX
    Explanations

    descriptions of locations

    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
    640
    +0.15
    0.7%
    454
    +0.13
    0.6%
    430
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.15
    0.10
    454
    +0.13
    0.08
    440
    +0.12
    0.07
    Negative Logits
    omics
    -0.80
    uly
    -0.79
     aven
    -0.75
     correct
    -0.73
     renegoti
    -0.72
     consistency
    -0.71
     rescind
    -0.70
     affili
    -0.70
     pardon
    -0.69
     ally
    -0.69
    POSITIVE LOGITS
     Dozens
    1.39
     Visitors
    1.38
     Located
    1.34
     Occasionally
    1.30
     Residents
    1.28
     Each
    1.23
     Hundreds
    1.19
     Suddenly
    1.18
     Behind
    1.15
     Thousands
    1.15
    Activations Density 0.360%

    No Known Activations