© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 7234
    Prev
    Next
    INDEX
    Explanations

    phrases related to technology or software

    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
    408
    +0.15
    0.7%
    361
    +0.11
    0.5%
    16
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    16
    +0.15
    0.04
    361
    +0.11
    0.04
    608
    +0.10
    0.05
    Negative Logits
    rocal
    -0.99
    terness
    -0.82
    ancies
    -0.81
    ouble
    -0.81
    unctions
    -0.78
     Cosponsors
    -0.77
    resy
    -0.76
    Å¡
    -0.76
    arre
    -0.75
    redible
    -0.74
    POSITIVE LOGITS
     EVER
    1.04
     ever
    1.00
     imaginable
    0.95
     mankind
    0.87
     humankind
    0.85
    ever
    0.77
     whistlebl
    0.72
     Ever
    0.69
     Mankind
    0.67
     maximizing
    0.66
    Activations Density 0.287%

    No Known Activations