© 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. 128194
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on numeric tokens (integers and decimals) in code and documentation.

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

    results and reports

    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
     a
    -1.11
    Honduras
    -0.91
    <?
    -0.91
    man
    -0.89
     so
    -0.89
    伯爵
    -0.88
    aky
    -0.88
     привет
    -0.88
     United
    -0.88
     can
    -0.87
    POSITIVE LOGITS
     results
    1.22
     bruja
    1.13
     maaf
    1.13
     ORDERED
    1.08
     salidas
    1.08
     rapor
    1.07
     demokra
    1.06
     udang
    1.06
     kampus
    1.05
     INDIC
    1.04
    Activations Density 0.022%

    No Known Activations