© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 39493
    Prev
    Next
    INDEX
    Explanations

    code delimiters and symbols

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

    The neuron fires on numeric literals—i.e. number tokens (integers or floats) embedded 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_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
     営業
    -0.89
     FirebaseAuth
    -0.89
     متح
    -0.85
     جائ
    -0.85
    那些
    -0.84
    нему
    -0.84
    يب
    -0.82
    气
    -0.81
     बेह
    -0.81
    Months
    -0.80
    POSITIVE LOGITS
     without
    1.04
     tropicales
    1.00
     béisbol
    0.98
     through
    0.96
     utilizando
    0.95
     acento
    0.91
     on
    0.91
     before
    0.90
    одо
    0.89
    ѭ
    0.89
    Activations Density 0.029%

    No Known Activations