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

    Neuronpedia

    Jacobian LensNEW
    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. 12871
    Prev
    Next
    INDEX
    Explanations

    text related to software development and technical details, including code snippets and explanations

    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
    447
    +0.19
    0.9%
    87
    +0.13
    0.6%
    727
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    267
    +0.19
    0.06
    183
    +0.13
    0.05
    50
    +0.10
    0.05
    Negative Logits
    .","
    -0.87
    .</
    -0.74
     ..."
    -0.64
     ÂŃ
    -0.64
     ''
    -0.63
    ='
    -0.63
     swe
    -0.63
     âĢķ
    -0.62
     �
    -0.61
     ``(
    -0.59
    POSITIVE LOGITS
     Conclusion
    1.35
    resa
    1.06
    odore
    1.03
    notations
    1.01
     Solution
    0.94
    endix
    0.94
     Lastly
    0.93
    mosp
    0.93
    wards
    0.90
    withstanding
    0.90
    Activations Density 0.449%

    No Known Activations