© 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. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 78833
    Prev
    Next
    INDEX
    Explanations

    Traceback errors

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron fires strongly on tokens that mark comment‐header metadata (e.g. license “BEGIN/​BLOCK,” origin paths, copyright notices, “jquery” comments), i.e. it detects comment lines that introduce file or license information.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     bismuth
    -0.75
     oracles
    -0.73
    Catalogue
    -0.72
     clemen
    -0.71
     Katalog
    -0.69
     մի
    -0.69
     IOT
    -0.69
     Husband
    -0.69
    オフィス
    -0.68
    gregation
    -0.68
    POSITIVE LOGITS
     exporting
    0.89
    "](
    0.86
    coring
    0.84
    itten
    0.81
    웠
    0.80
     noche
    0.80
    dited
    0.79
    🢒
    0.79
    "]}
    0.79
    '],
    0.79
    Activations Density 0.001%

    No Known Activations