© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 116408
    Prev
    Next
    INDEX
    Explanations

    version numbers and identifiers

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

    The neuron is detecting structured “key” tokens—like metadata field names or language/locale tags (e.g. de_DE, AT, description, version, pH)—that appear at the start of lines in front-matter or configuration blocks.

    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
     langkah
    -0.85
    muñ
    -0.77
     there
    -0.77
     secured
    -0.76
     daar
    -0.75
     honlap
    -0.75
    dari
    -0.74
    chg
    -0.74
    部の
    -0.72
    ברית
    -0.72
    POSITIVE LOGITS
     lizard
    0.75
     общественного
    0.74
    Parte
    0.73
     передает
    0.73
    Proces
    0.73
     сообщает
    0.72
     クロス
    0.72
    Fiz
    0.71
    Fut
    0.71
     pope
    0.71
    Activations Density 0.010%

    No Known Activations