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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 14289
    Prev
    Next
    INDEX
    Explanations

    driving sales, penalties, legacy

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

    The main thing this neuron does is pick out rare or out-of-vocabulary tokens—essentially stray metadata or float-artifact tokens rather than normal content 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
    string
    0.88
    function
    0.81
    [(
    0.80
    [][
    0.78
    >)
    0.76
    "
    0.76
    "_
    0.75
    >
    0.73
    세대
    0.73
    for
    0.73
    POSITIVE LOGITS
     toate
    0.80
     FirebaseAuth
    0.80
     semua
    0.77
     Django
    0.77
     persever
    0.76
     embal
    0.76
     Musik
    0.75
     Tiktok
    0.75
     django
    0.74
     monastic
    0.71
    Activations Density 0.002%

    No Known Activations