© 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. 92
    Prev
    Next
    INDEX
    Explanations

    phrases related to stories or situations

    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
    +0.26
    1.2%
    480
    +0.19
    0.9%
    679
    +0.14
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    552
    +0.26
    0.03
    655
    +0.19
    0.03
    673
    +0.14
    0.02
    Negative Logits
    bledon
    -0.61
     Examiner
    -0.60
     Seym
    -0.58
    erenn
    -0.57
     Borders
    -0.57
     Frie
    -0.56
    Ire
    -0.54
     Depot
    -0.54
    Adams
    -0.54
     Gibraltar
    -0.53
    POSITIVE LOGITS
     violates
    0.85
     lasted
    0.78
     doesnt
    0.76
     translates
    0.75
     consists
    0.75
     includes
    0.73
     resembles
    0.73
     utilizes
    0.72
     accompanies
    0.72
     consisted
    0.71
    Activations Density 0.055%

    No Known Activations