© 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. 14267
    Prev
    Next
    INDEX
    Explanations

    the word "yet" as a significant pattern

    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
    671
    +0.10
    0.5%
    570
    +0.09
    0.4%
    108
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    671
    +0.10
    0.02
    221
    +0.09
    0.02
    38
    +0.09
    0.02
    Negative Logits
    gencies
    -0.69
    little
    -0.67
     teasp
    -0.64
    ranged
    -0.62
    Ĥª
    -0.62
    ership
    -0.61
    lua
    -0.61
    rather
    -0.60
     RTX
    -0.59
    pill
    -0.58
    POSITIVE LOGITS
    hin
    0.94
    icable
    0.93
     formally
    0.89
     officially
    0.89
     finalized
    0.82
     reconcil
    0.82
     realized
    0.78
     reached
    0.78
     fully
    0.74
     surpassed
    0.74
    Activations Density 0.024%

    No Known Activations