© 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 1536
    9. 8-RES_FS1536-JB
    11. 1068
    Prev
    Next
    INDEX
    Explanations

    superlatives or things that are the best

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    450
    +0.10
    0.5%
    408
    +0.10
    0.4%
    296
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    422
    +0.10
    0.08
    299
    +0.10
    0.09
    16
    +0.10
    0.07
    Negative Logits
    rocal
    -0.79
    actionDate
    -0.77
    Query
    -0.71
     Cosponsors
    -0.68
    ãĥĥ
    -0.68
    Å¡
    -0.67
    çīĪ
    -0.66
    ead
    -0.66
    affle
    -0.65
     unfocusedRange
    -0.63
    POSITIVE LOGITS
     EVER
    1.36
     ever
    1.26
     imaginable
    1.12
    ever
    0.99
     anywhere
    0.96
     humankind
    0.93
     Ever
    0.91
     mankind
    0.87
     shortest
    0.81
    Ever
    0.77
    Activations Density 2.455%

    No Known Activations