© 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. 13822
    Prev
    Next
    INDEX
    Explanations

    transitive verbs denoting a specific action

    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
    335
    +0.14
    0.7%
    166
    +0.11
    0.5%
    327
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    335
    +0.14
    0.05
    41
    +0.11
    0.04
    298
    +0.10
    0.04
    Negative Logits
    relations
    -0.79
    PN
    -0.74
    Write
    -0.74
    ochemistry
    -0.73
    gans
    -0.73
    arians
    -0.71
    MQ
    -0.70
    oci
    -0.69
    actionDate
    -0.69
    Ware
    -0.68
    POSITIVE LOGITS
     cliffs
    1.50
     heights
    1.42
     tallest
    1.33
     ceilings
    1.33
     ceiling
    1.28
     towering
    1.26
     stairs
    1.23
     cliff
    1.22
     skysc
    1.22
     towers
    1.18
    Activations Density 0.416%

    No Known Activations