© 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-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 353
    Prev
    Next
    INDEX
    Explanations

    journal article

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

    The neuron is detecting numeric tokens (e.g. years, measurements, decimals, and other numbers).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    
    0.52
     emble
    0.49
     sizlere
    0.48
     odio
    0.48
     jouw
    0.47
     monstru
    0.47
     supernat
    0.46
     sizlerle
    0.46
     väldigt
    0.46
     dolayı
    0.46
    POSITIVE LOGITS
    P
    0.62
    S
    0.59
    U
    0.59
    R
    0.56
    In
    0.55
    A
    0.55
    B
    0.55
    M
    0.54
    L
    0.53
    F
    0.52
    Activations Density 0.441%

    No Known Activations