© 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. 10-GEMMASCOPE-RES-131K
    7. 86842
    Prev
    Next
    INDEX
    Explanations

    school pupils

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

    The neuron strongly activates on numeric count tokens—especially figures giving numbers of pupils—i.e. it finds numeric statistics.

    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
    ↵
    -3.84
    颙
    -2.77
     don
    -2.69
    ”
    -2.45
    呖
    -2.44
     become
    -2.38
     inexplicable
    -2.31
    硨
    -2.27
    溲
    -2.25
     caneca
    -2.17
    POSITIVE LOGITS
     も
    2.69
    塤
    2.67
    when
    2.44
    their
    2.31
     accordance
    2.31
    皷
    2.30
    厹
    2.28
    蹌
    2.25
     OGSÅ
    2.23
     Also
    2.17
    Activations Density 0.003%

    No Known Activations