© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 4551
    Prev
    Next
    INDEX
    Explanations

    tables, rows, and csv

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

    The neuron is keyed to tokens marking document‐level references or pointers—things like URLs, feed or citation markers, and similar “metadata” tokens.

    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
    ۃ
    0.79
    }($
    0.78
    νος
    0.76
    routingHeader
    0.73
    อส
    0.73
     λόγω
    0.73
    ণ
    0.71
    )"))
    0.71
    <unused2108>
    0.70
    Sight
    0.70
    POSITIVE LOGITS
    table
    0.76
     glimmer
    0.73
     improvis
    0.72
     boilerplate
    0.71
     unnecessary
    0.71
     prompts
    0.70
     कवाय
    0.70
     summarized
    0.69
    bleau
    0.68
     improvised
    0.68
    Activations Density 0.004%

    No Known Activations