© 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. 12585
    Prev
    Next
    INDEX
    Explanations

    technical terms or codes

    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
    378
    +0.16
    0.7%
    87
    +0.12
    0.5%
    373
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.16
    0.03
    130
    +0.12
    0.02
    674
    +0.11
    0.03
    Negative Logits
    packing
    -0.60
    ement
    -0.59
     conv
    -0.59
     cle
    -0.57
    hom
    -0.55
     manifest
    -0.54
     Kot
    -0.53
     gent
    -0.53
     grounds
    -0.50
     manifests
    -0.50
    POSITIVE LOGITS
    "]=>
    0.82
    ª
    0.73
    thora
    0.71
    ļéĨĴ
    0.71
     horm
    0.67
    ================================================================
    0.66
    Ability
    0.66
    ESSION
    0.66
    Ļ½
    0.66
    ij
    0.66
    Activations Density 0.128%

    No Known Activations