© 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. 129020
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on numeric tokens and punctuation associated with numbers, effectively detecting numeric expressions in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    list item delimiters

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     what
    -1.23
    ditionally
    -1.19
     these
    -1.18
     thing
    -1.00
     this
    -0.99
     These
    -0.96
     how
    -0.96
     get
    -0.95
     Please
    -0.95
    licante
    -0.94
    POSITIVE LOGITS
     hemd
    1.05
    ſes
    1.04
     pengar
    1.03
     konfig
    1.01
    erreur
    0.98
     karte
    0.98
    solete
    0.96
     การ
    0.95
    Handlung
    0.95
    joueurs
    0.94
    Activations Density 0.028%

    No Known Activations