© 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 EvalsExportsinterp-engineNEWAPI 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. 2529
    Prev
    Next
    INDEX
    Explanations

    references to quantity, specifically the word "many" followed by a number

    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
    47
    +0.11
    0.5%
    158
    +0.10
    0.5%
    320
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    654
    +0.11
    0.02
    252
    +0.10
    0.02
    408
    +0.10
    0.02
    Negative Logits
    Ľ
    -0.78
    ĺħ
    -0.76
     Butterfly
    -0.71
    ardo
    -0.68
     Brav
    -0.67
    strong
    -0.63
    NetMessage
    -0.63
    agher
    -0.62
    ĪĴ
    -0.61
    chieve
    -0.60
    POSITIVE LOGITS
     digits
    0.69
    theless
    0.67
     Than
    0.65
    awi
    0.64
     AS
    0.63
     than
    0.62
    antz
    0.61
     copies
    0.60
     practicable
    0.59
    enery
    0.58
    Activations Density 0.030%

    No Known Activations