© 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. 31-GEMMASCOPE-2-RES-16K
    4. 15641
    Prev
    Next
    INDEX
    Explanations

    `pd.to_datetime`

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

    The neuron detects call-to-action directives—phrases that prompt the reader to click, view, send, or otherwise take an action.

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

    calls to action in instructional or interactive contexts, particularly phrases inviting the reader to perform an action like "click to read more," "to share," "to send," or "to view."

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

    instructions and calls to action, especially infinitive “to + verb” directives common in links and web prompts.

    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_31_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
     וב
    0.58
    来进行
    0.56
     পাকিস্তানকে
    0.55
    urgical
    0.54
     synthèse
    0.53
     kepada
    0.52
    ovich
    0.52
    đena
    0.52
     come
    0.52
     إلى
    0.52
    POSITIVE LOGITS
     extent
    0.65
     ול
    0.63
    arno
    0.61
    ibir
    0.59
    extent
    0.58
    やつ
    0.54
     çıkan
    0.54
    tener
    0.54
     iniciar
    0.53
     خلق
    0.53
    Activations Density 0.129%

    No Known Activations