© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 45211
    Prev
    Next
    INDEX
    Explanations

    proper nouns and abbreviations

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

    The neuron fires most strongly on isolated non-standard characters and symbols—e.g. punctuation like “|” or extended‐Latin letters with diacritics (ď)—and also (more weakly) on standalone place names, so overall it’s detecting unusual or non-alphanumeric tokens.

    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
    ्य
    -2.55
     ſch
    -2.52
    的吧
    -2.34
    的很
    -2.28
    客様
    -2.25
    等は
    -2.19
    ing
    -2.17
    ‟
    -2.14
     ſou
    -2.14
    forKey
    -2.14
    POSITIVE LOGITS
    ,’
    2.78
    You
    2.78
    If
    2.75
    ve
    2.69
    We
    2.64
    Our
    2.61
    ’
    2.58
    ,”
    2.53
    These
    2.53
    Maybe
    2.52
    Activations Density 0.000%

    No Known Activations