© 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. 6117
    Prev
    Next
    INDEX
    Explanations

    legal or criminal terms related to law enforcement and animal welfare issues

    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
    36
    +0.30
    1.5%
    393
    +0.17
    0.9%
    314
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.30
    0.22
    250
    +0.17
    0.10
    636
    +0.12
    0.09
    Negative Logits
    ESPN
    -0.73
    grounds
    -0.72
    Links
    -0.70
    Enjoy
    -0.69
    Examples
    -0.69
     examples
    -0.68
    Rules
    -0.68
    tests
    -0.68
    !!!!!
    -0.68
    Ub
    -0.66
    POSITIVE LOGITS
    versive
    1.04
    rouse
    0.98
     lot
    0.91
     particular
    0.91
     substantial
    0.89
     predetermined
    0.88
     multitude
    0.87
     significant
    0.87
     favorable
    0.86
     hefty
    0.86
    Activations Density 0.536%

    No Known Activations