© 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. 13605
    Prev
    Next
    INDEX
    Explanations

    error messages related to programming code

    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
    281
    +0.11
    0.5%
    619
    +0.10
    0.5%
    757
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.11
    0.03
    551
    +0.10
    0.03
    139
    +0.10
    0.03
    Negative Logits
    prof
    -0.66
     Spectre
    -0.64
    pid
    -0.63
    PDATE
    -0.62
    tra
    -0.62
    phase
    -0.61
    fram
    -0.61
    mission
    -0.59
    sense
    -0.59
     threaded
    -0.58
    POSITIVE LOGITS
     be
    1.28
     BE
    1.00
    BE
    0.97
    Be
    0.96
    be
    0.94
     been
    0.90
     Be
    0.86
     being
    0.81
     become
    0.80
    ãĥĻ
    0.79
    Activations Density 0.132%

    No Known Activations