© 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. 22-GEMMASCOPE-RES-131K
    7. 66144
    Prev
    Next
    INDEX
    Explanations

    following first words"

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

    The neuron fires strongly on numeric tokens (e.g. version numbers, list indices, dates) throughout the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     δεν
    -1.34
     calendriers
    -1.27
    сле
    -1.27
     Perubahan
    -1.23
    idyl
    -1.17
     Offisielt
    -1.16
    vaal
    -1.14
    kval
    -1.14
     Vainqueur
    -1.14
     ikke
    -1.12
    POSITIVE LOGITS
     therefore
    1.23
    やっぱ
    1.20
     basically
    1.15
    ほんとに
    1.14
    多么
    1.02
     haupts
    1.02
    これからも
    1.00
    顿时
    1.00
     wichtigen
    1.00
     soorten
    0.97
    Activations Density 0.174%

    No Known Activations