© 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. 75225
    Prev
    Next
    INDEX
    Explanations

    addition

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

    The neuron fires primarily on numeric tokens (digits, percentages, timestamps, code‐related numbers), i.e. it detects numbers.

    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
    <0xA1>
    -0.82
    uksessa
    -0.81
     posteriormente
    -0.78
     następnie
    -0.77
     jednot
    -0.76
    othelioma
    -0.75
    SneakyThrows
    -0.75
    inique
    -0.73
     overflowing
    -0.72
    toArray
    -0.71
    POSITIVE LOGITS
     추가
    1.09
     addon
    1.05
     дополнитель
    0.99
     aggiunto
    0.98
     additional
    0.95
     addons
    0.93
     tambahan
    0.93
     complemento
    0.93
     addition
    0.91
     zusätzliche
    0.90
    Activations Density 0.115%

    No Known Activations