© 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. 39506
    Prev
    Next
    INDEX
    Explanations

    numbers less than one

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

    The neuron fires on numeric literals (integers or decimals) wherever they appear.

    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
     will
    -1.86
     =
    -1.70
     allows
    -1.58
     gives
    -1.57
     makes
    -1.53
     but
    -1.42
     provides
    -1.42
     would
    -1.42
     ensures
    -1.41
    ставляет
    -1.39
    POSITIVE LOGITS
     rengi
    1.53
    嘥
    1.49
    这些
    1.48
    ⸫
    1.45
     désir
    1.45
    useful
    1.43
     fameux
    1.38
     he
    1.38
    多かった
    1.37
     różnych
    1.34
    Activations Density 0.016%

    No Known Activations