© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3227
    Prev
    Next
    INDEX
    Explanations

    numbers and quantity

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

    The neuron fires strongly on standalone numeric tokens (e.g. page or item 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_24_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
    生ま
    0.45
    部に
    0.42
    монт
    0.41
    િ
    0.41
     বৈশিষ্ট
    0.41
    ೇ
    0.41
    一代
    0.39
    риб
    0.39
    ើង
    0.39
     Trời
    0.38
    POSITIVE LOGITS
     cheat
    0.43
     cheats
    0.43
     leçon
    0.40
     cheating
    0.40
     I
    0.39
     _____
    0.39
     October
    0.38
     şu
    0.38
     Jane
    0.38
     isothermal
    0.38
    Activations Density 0.001%

    No Known Activations