© 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. 25683
    Prev
    Next
    INDEX
    Explanations

    mathematical/technical identifiers

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

    The neuron is triggered by explicit character‐class listings inside square brackets—e.g. the comma‐free lists of letters or digits you see in regex-style “[…]” expressions.

    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
     would
    -0.85
     until
    -0.75
    リューション
    -0.73
     Willing
    -0.71
    äts
    -0.69
    再见
    -0.69
    ropho
    -0.69
     encro
    -0.69
    不肯
    -0.69
     в
    -0.68
    POSITIVE LOGITS
    stw
    0.83
     balkon
    0.82
     mily
    0.82
     dezelve
    0.82
     setw
    0.82
     KLM
    0.82
     Mungkin
    0.81
     restau
    0.81
    KOS
    0.80
    ctivité
    0.80
    Activations Density 0.045%

    No Known Activations