© 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. 8582
    Prev
    Next
    INDEX
    Explanations

    numeric values and personal identifiers in a text related to digital identity and security

    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
    556
    +0.11
    0.5%
    252
    +0.10
    0.4%
    639
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    639
    +0.11
    0.03
    393
    +0.10
    0.04
    384
    +0.09
    0.02
    Negative Logits
    roo
    -0.76
    iculture
    -0.71
    fires
    -0.71
    raq
    -0.70
    conservative
    -0.69
    geons
    -0.68
    ocrats
    -0.67
    artisan
    -0.67
    irit
    -0.67
    arten
    -0.66
    POSITIVE LOGITS
     initials
    1.22
     address
    1.16
     Address
    1.15
     ZIP
    1.11
     surname
    1.11
     nationality
    1.11
     location
    1.10
     addresses
    1.08
     coordinates
    1.06
    estamp
    1.05
    Activations Density 0.239%

    No Known Activations