© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 13261
    Prev
    Next
    INDEX
    Explanations

    URLs and examples

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

    This neuron specializes in detecting news‐style datelines, i.e. city or location names typically followed by a comma at the start of an article.

    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
    बॉलीवुड
    0.39
    시대
    0.38
    الحمد
    0.38
    фесси
    0.37
    ޙ
    0.37
    <unused78>
    0.36
     ګرځ
    0.36
    гун
    0.35
    koliko
    0.34
    <unused21>
    0.34
    POSITIVE LOGITS
     точки
    0.38
    िस
    0.37
     pinturas
    0.36
     యొక్క
    0.35
     three
    0.34
    ände
    0.34
     gebruikt
    0.34
     生成
    0.33
     B
    0.32
    のが
    0.32
    Activations Density 0.004%

    No Known Activations