© 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. 67370
    Prev
    Next
    INDEX
    Explanations

    programming statement endings

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

    The neuron strongly activates on numeric literals or tokens containing digits.

    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
     desigual
    -0.96
    ismatic
    -0.92
    described
    -0.88
    肯定
    -0.85
    mentaire
    -0.85
     termos
    -0.85
    styleUrls
    -0.84
    licopter
    -0.83
    ContentAsync
    -0.82
    ющим
    -0.80
    POSITIVE LOGITS
     escalier
    0.81
    隴
    0.80
     even
    0.78
    EDWARD
    0.77
     ecosystems
    0.76
     modelleri
    0.76
     弁当
    0.75
     тоже
    0.75
     täglichen
    0.75
    
    0.74
    Activations Density 0.006%

    No Known Activations