© 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. 10-GEMMASCOPE-RES-131K
    7. 67503
    Prev
    Next
    INDEX
    Explanations

    contractions and specific punctuation

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

    The neuron fires on numeric tokens (particularly numbers with digits/decimals, e.g. version numbers, prices, measurements).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     it
    -1.58
     Französ
    -1.51
     and
    -1.49
     each
    -1.41
    東南
    -1.35
    党的
    -1.30
     kathol
    -1.29
    明確
    -1.28
     Gemeinder
    -1.27
     katholischen
    -1.27
    POSITIVE LOGITS
    fficking
    1.45
     Lê
    1.43
    ਗ਼
    1.38
    出版年
    1.36
     préstamos
    1.34
    foundland
    1.33
    urent
    1.32
    érape
    1.30
     ofrecemos
    1.30
     manquer
    1.29
    Activations Density 0.423%

    No Known Activations