© 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. 13683
    Prev
    Next
    INDEX
    Explanations

    association agreement

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

    The neuron isn’t picking out any coherent semantic class but rather fires strongly on a handful of specific token identities (e.g. “find,” “horses,” “I’m,” “there’s,” etc.), effectively acting like a lookup for those particular words.

    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
    ور
    0.88
    ٦
    0.88
    ВО
    0.87
    Χ
    0.82
     elution
    0.82
    ﺩ
    0.82
    ٥
    0.80
    كز
    0.77
     batsmen
    0.77
    ۱۳
    0.77
    POSITIVE LOGITS
    tea
    0.99
    law
    0.91
    aud
    0.90
    mar
    0.89
    s
    0.88
    duc
    0.87
    sing
    0.87
    t
    0.87
    sa
    0.86
    dec
    0.86
    Activations Density 0.000%

    No Known Activations