© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 1945
    Prev
    Next
    INDEX
    Explanations

    arrival/departure

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

    This neuron fires strongly on unusually rare or low‐frequency tokens (i.e. high‐IDF words).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    low-quality or spam web content, particularly from commercial or scraped websites.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    common connective and structural words in informational prose—prepositions, determiners, and general modifiers that link, qualify, or quantify statements.

    oai_token-act-pair · gpt-5Triggered 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
    luster
    0.80
    هل
    0.80
    ebe
    0.71
    kontakt
    0.71
    lasma
    0.70
    Tabla
    0.70
    ال
    0.69
    اؤ
    0.69
    ar
    0.68
     以上
    0.68
    POSITIVE LOGITS
    щего
    0.89
    केक
    0.80
     distância
    0.76
     Sutherland
    0.76
     Wynn
    0.75
     Gonzaga
    0.75
    蹴
    0.75
    ချ
    0.75
     Wach
    0.75
    तियां
    0.75
    Activations Density 0.000%

    No Known Activations