© 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. 59761
    Prev
    Next
    INDEX
    Explanations

    code definitions and structures

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

    This neuron is primarily activated by all-uppercase identifiers (e.g. macro names, constant names, and other all-caps tokens) in code.

    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
     aslında
    -0.89
    MORE
    -0.81
    лік
    -0.79
    ys
    -0.78
     村
    -0.76
    ło
    -0.75
    add
    -0.74
    ifax
    -0.73
     once
    -0.73
       
    -0.72
    POSITIVE LOGITS
     диви
    0.86
     because
    0.81
    STILL
    0.81
    鶯
    0.80
     weil
    0.80
    ::_
    0.78
    出来ました
    0.78
    Datuak
    0.78
    道を
    0.77
    Estad
    0.77
    Activations Density 0.007%

    No Known Activations