© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 73598
    Prev
    Next
    INDEX
    Explanations

    programming syntax and structures

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

    The main thing this neuron does is find numeric literals (numbers) in the text.

    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
     from
    -1.77
    摆手
    -1.71
     that
    -1.71
     PUSTAKA
    -1.66
     what
    -1.62
     by
    -1.52
    歿
    -1.49
     for
    -1.45
     know
    -1.37
     Like
    -1.36
    POSITIVE LOGITS
    9
    1.73
    te
    1.70
    6
    1.67
     inga
    1.66
    ’
    1.62
    5
    1.58
     henne
    1.58
     tilla
    1.57
     deras
    1.53
     lille
    1.51
    Activations Density 0.048%

    No Known Activations