© 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. 24541
    Prev
    Next
    INDEX
    Explanations

    phrases indicating advice or recommendations

    oai_token-act-pair · gpt-4o-miniTriggered 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
    620
    +0.11
    0.5%
    608
    +0.11
    0.5%
    88
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.11
    0.05
    88
    +0.11
    0.04
    22
    +0.11
    0.04
    Negative Logits
    chenko
    -0.76
    igger
    -0.70
    ouble
    -0.68
    hya
    -0.67
     Accessed
    -0.65
    zona
    -0.64
    ohn
    -0.63
    querque
    -0.63
    ãĤ´ãĥ³
    -0.63
    /.
    -0.62
    POSITIVE LOGITS
     starters
    1.11
    bidden
    1.09
     purposes
    1.08
     beginners
    1.08
    geries
    1.07
     reasons
    0.99
     everyone
    0.94
     outsiders
    0.94
     anyone
    0.92
     everybody
    0.92
    Activations Density 0.129%

    No Known Activations