© 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. 79646
    Prev
    Next
    INDEX
    Explanations

    names, headings, and specific terms

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

    This neuron primarily detects numeric tokens—digits or number literals—wherever they appear 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
     both
    -2.75
     without
    -2.66
     or
    -2.44
     with
    -2.41
     in
    -2.31
     very
    -2.31
     before
    -2.27
     as
    -2.27
     on
    -2.23
     when
    -2.19
    POSITIVE LOGITS
     jezt
    2.09
     gigantic
    2.00
     Dinas
    1.99
     horrid
    1.98
     insanely
    1.91
     astounding
    1.86
     ludicrous
    1.80
     MAKES
    1.80
     スリーブ
    1.80
     ridiculously
    1.77
    Activations Density 0.005%

    No Known Activations