© 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. 16612
    Prev
    Next
    INDEX
    Explanations

    conjunctions or introductory phrases indicating contrast or alternative information

    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
    483
    +0.13
    0.6%
    115
    +0.11
    0.5%
    565
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    483
    +0.13
    0.05
    208
    +0.11
    0.05
    8
    +0.11
    0.06
    Negative Logits
    QB
    -0.80
    ãĤ©
    -0.73
    é¾
    -0.68
    ãĥĺ
    -0.63
    Nation
    -0.63
    Border
    -0.62
    çľ
    -0.62
    Frames
    -0.59
    ãĥĩ
    -0.59
    Unit
    -0.59
    POSITIVE LOGITS
     nevertheless
    1.27
     nonetheless
    1.12
     anecd
    0.86
    efficients
    0.71
     occasionally
    0.70
    netflix
    0.69
    ulia
    0.67
    ebus
    0.65
    poons
    0.64
     Cathy
    0.64
    Activations Density 0.397%

    No Known Activations