© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100578
    Prev
    Next
    INDEX
    Explanations

    counting digits

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

    the digit "3" appearing in numeric sequences, particularly in code, addresses, dates, and version numbers.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    This neuron detects digit characters within numeric strings (especially the digit “3”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    icksburg
    -1.29
    álisis
    -1.23
    versies
    -1.21
    囯
    -1.20
    événement
    -1.20
    tadores
    -1.20
     caminhão
    -1.18
     prêmio
    -1.18
     OGSÅ
    -1.15
    ører
    -1.14
    POSITIVE LOGITS
    I
    1.40
     musíte
    1.17
    rillo
    1.09
    俠
    1.09
     женская
    1.09
    ка
    1.08
     étrange
    1.05
     Observations
    1.05
     ботинки
    1.05
    3
    1.04
    Activations Density 0.041%

    No Known Activations