© 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. 989
    Prev
    Next
    INDEX
    Explanations

    Vector Network

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

    The neuron is picking out numeric tokens, especially multi‐digit and floating‐point 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
     Área
    0.92
     Audience
    0.87
    Appl
    0.87
     Enseñanza
    0.84
     Nutzer
    0.82
    U
    0.80
    jed
    0.79
     Adaptive
    0.79
     జ్ఞ
    0.79
     regionales
    0.77
    POSITIVE LOGITS
     conquered
    0.76
     carène
    0.69
     tormented
    0.69
     catapult
    0.66
    וה
    0.64
    ᱼ
    0.63
    犸
    0.63
     swayed
    0.63
     arred
    0.62
     wept
    0.62
    Activations Density 0.075%

    No Known Activations