© 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. 93219
    Prev
    Next
    INDEX
    Explanations

    code constructs with parentheses

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

    The neuron fires on programming‐style tokens—especially code punctuation (e.g. “<” and “>” in generics) and numeric literals—i.e. it identifies snippets of source 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_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
    ledad
    -1.07
     anunció
    -1.07
    entown
    -1.02
    uminum
    -1.01
    termilk
    -1.00
     no
    -0.98
    ternut
    -0.96
    Abraços
    -0.92
    prende
    -0.90
     guapos
    -0.88
    POSITIVE LOGITS
    că
    1.04
    ēl
    1.03
     nevo
    1.00
     ваше
    0.98
    olá
    0.98
    arono
    0.97
    itária
    0.96
     또
    0.95
     Slough
    0.93
     допомогою
    0.92
    Activations Density 0.003%

    No Known Activations