© 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. 39720
    Prev
    Next
    INDEX
    Explanations

    data types and identifiers

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

    The main thing this neuron does is detect numeric literal tokens (e.g. version IDs, floating-point numbers, integer constants) in source code.

    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
     合集
    -1.73
    鋮
    -1.70
    蹓
    -1.67
     vixion
    -1.63
     fakta
    -1.63
     スタン
    -1.60
     cruda
    -1.60
    Selanjutnya
    -1.59
    鼉
    -1.56
    屨
    -1.56
    POSITIVE LOGITS
     of
    2.88
     to
    1.57
     her
    1.45
     keine
    1.42
    !
    1.36
     ma
    1.36
     off
    1.34
    榴
    1.34
     on
    1.34
     '
    1.30
    Activations Density 0.001%

    No Known Activations