© 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. 34184
    Prev
    Next
    INDEX
    Explanations

    foreign language snippets

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

    The neuron fires on the key concrete nouns naming what is being counted or reported (e.g. ammunition, stores, power stations, penalties, birds, dollars) in quantified news contexts.

    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
     Although
    -1.89
     Often
    -1.82
     With
    -1.78
     While
    -1.77
     really
    -1.77
     extraordinarily
    -1.74
     Despite
    -1.74
     as
    -1.73
     After
    -1.70
     unrelenting
    -1.70
    POSITIVE LOGITS
     dando
    1.82
     Dinas
    1.77
    などが
    1.77
     և
    1.77
     faciliter
    1.76
     zoude
    1.75
    ebly
    1.71
     alguns
    1.70
     durante
    1.70
    ᕮ
    1.67
    Activations Density 0.126%

    No Known Activations