© 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. 4837
    Prev
    Next
    INDEX
    Explanations

    phrases related to forceful movement or pressure

    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
    501
    +0.11
    0.5%
    540
    +0.10
    0.5%
    178
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.02
    526
    +0.10
    0.00
    501
    +0.10
    0.01
    Negative Logits
    keeper
    -0.87
    externalActionCode
    -0.86
    keepers
    -0.70
    redits
    -0.68
     CLSID
    -0.66
    ãĤ¿
    -0.64
    cius
    -0.63
    displayText
    -0.63
    enfranch
    -0.63
    keeping
    -0.61
    POSITIVE LOGITS
    ing
    0.86
    sburg
    0.83
    igating
    0.81
     thrust
    0.78
    ting
    0.78
    onite
    0.76
    ments
    0.76
    icularly
    0.75
    anche
    0.75
    nesday
    0.74
    Activations Density 0.038%

    No Known Activations