© 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. 8301
    Prev
    Next
    INDEX
    Explanations

    . , 0 , 3 , X , right , taller , write

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

    The neuron is keyed to “content” tokens—that is, it strongly activates on substantive words (nouns, main verbs, names, etc.) rather than on common function or filler 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
    ఙ
    0.94
    णी
    0.80
    有可能
    0.80
    ন
    0.80
    nof
    0.75
    ến
    0.73
    דר
    0.72
    ේෂ
    0.71
     বিক্ষোভ
    0.71
    nant
    0.70
    POSITIVE LOGITS
     ette
    0.78
     Этот
    0.75
     этот
    0.75
     эту
    0.72
    ቱን
    0.71
    Postal
    0.70
     ブリヂストン
    0.69
    BURGH
    0.68
    メカ
    0.68
     Это
    0.67
    Activations Density 0.000%

    No Known Activations