© 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. 34-GEMMASCOPE-RES-131K
    7. 82769
    Prev
    Next
    INDEX
    Explanations

    code termination characters

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

    The neuron fires on numeric literals (sequences of digits) in code, regardless of context or language.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    揽
    -1.00
    ПЕ
    -0.90
     need
    -0.89
     besondere
    -0.89
     نکن
    -0.87
     plaintiffs
    -0.86
    ほら
    -0.85
    णा
    -0.85
     toit
    -0.82
     vær
    -0.82
    POSITIVE LOGITS
     because
    1.06
     still
    0.91
     песни
    0.82
     each
    0.82
     Вя
    0.81
    because
    0.79
    kosť
    0.75
    觉
    0.75
    əri
    0.75
    ↵
    0.75
    Activations Density 0.020%

    No Known Activations