© 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. 28049
    Prev
    Next
    INDEX
    Explanations

    mp4.1080

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

    The neuron fires most strongly on numeric literals (e.g. “0”, “32”, “143”, “208”, etc.)—i.e. tokens representing numbers.

    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
     Fakta
    -1.02
    agory
    -0.99
     rollercoaster
    -0.98
    mmä
    -0.98
    Viernes
    -0.98
     florales
    -0.94
     basit
    -0.93
    ながらも
    -0.91
     vont
    -0.90
     manipulations
    -0.90
    POSITIVE LOGITS
     that
    1.04
     somebody
    0.94
     By
    0.94
     secure
    0.93
     someone
    0.89
     akzep
    0.87
    選ば
    0.86
     ciò
    0.85
     нарушение
    0.84
     these
    0.84
    Activations Density 0.014%

    No Known Activations