© 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. 31-GEMMASCOPE-2-RES-16K
    7. 13184
    Prev
    Next
    INDEX
    Explanations

    number of terms, periods, or chunks

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

    The neuron primarily activates on numeric tokens (i.e. standalone 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
     Jair
    0.67
    漃
    0.58
     jude
    0.57
     multiplicar
    0.56
    ্ু
    0.56
     thisobject
    0.56
     criticize
    0.55
     interprets
    0.55
    ᅴ
    0.55
     तस्वीर
    0.54
    POSITIVE LOGITS
    Follow
    0.60
     Milk
    0.58
    Rus
    0.56
    K
    0.56
    Milk
    0.54
    데이터
    0.54
    MIL
    0.54
    Mag
    0.54
    Foi
    0.54
    Tobacco
    0.53
    Activations Density 0.000%

    No Known Activations