© 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 EvalsExportsinterp-engineNEWAPI 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. 2722
    Prev
    Next
    INDEX
    Explanations

    descriptions of events or actions with specific details such as names, dates, and 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
    436
    +0.11
    0.5%
    129
    +0.11
    0.5%
    373
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    295
    +0.11
    0.04
    415
    +0.11
    0.03
    582
    +0.10
    0.02
    Negative Logits
     Pastebin
    -0.84
     Repeat
    -0.71
    "]=>
    -0.69
    eworthy
    -0.69
    andowski
    -0.69
    Overall
    -0.68
    imgur
    -0.68
     Conclusion
    -0.67
    letes
    -0.67
     Cheong
    -0.65
    POSITIVE LOGITS
     fledgling
    1.08
     intrigued
    0.90
     nascent
    0.87
     burgeoning
    0.81
     promising
    0.80
     ambitious
    0.79
     dreamed
    0.78
     classmate
    0.76
     naïve
    0.76
     enroll
    0.75
    Activations Density 0.885%

    No Known Activations