© 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. 13091
    Prev
    Next
    INDEX
    Explanations

    some very

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

    The neuron fires on the key content word (usually a noun) that completes a common two-word phrase or collocation (e.g. “terms” in “in terms,” “toughness” in “fracture toughness,” “writing” in “in writing,” “heritage” in “heritage site,” etc.).

    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
    ومات
    0.76
    குடும்ப
    0.73
    اہر
    0.72
    戀
    0.71
    ስት
    0.71
    ุม
    0.70
    unjungi
    0.70
     অনুমোদন
    0.70
    біль
    0.69
    靂
    0.69
    POSITIVE LOGITS
     sporo
    0.88
    ía
    0.82
     a
    0.81
     пару
    0.81
     orange
    0.80
     sightings
    0.78
    ho
    0.76
     tingling
    0.76
     phosphorous
    0.76
     chom
    0.75
    Activations Density 0.000%

    No Known Activations