© 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. 128574
    Prev
    Next
    INDEX
    Explanations

    This neuron activates on numeric tokens (e.g. digit sequences and numbers).

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

    construction of

    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
     задний
    -1.20
     солнца
    -1.08
    })));
    -1.05
    ℂ
    -1.04
     որ
    -1.02
     pueden
    -1.01
    DECIMAL
    -1.01
    恣
    -1.01
     специальные
    -1.00
     высокие
    -0.99
    POSITIVE LOGITS
     philippe
    1.17
     وَ
    1.04
     paulo
    1.03
    瞥
    0.99
    Diction
    0.98
    parently
    0.96
    Ecole
    0.95
    sieurs
    0.95
     pernic
    0.94
     UserService
    0.93
    Activations Density 0.005%

    No Known Activations