© 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. 31-GEMMASCOPE-2-RES-16K
    7. 9867
    Prev
    Next
    INDEX
    Explanations

    can you / could you

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

    The neuron fires on modal auxiliary verbs (especially “can” and “will”) signaling questions or future‐tense statements.

    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_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
    ఇ
    0.66
    ганда
    0.63
     ő
    0.57
    пример
    0.56
     ok
    0.55
     quella
    0.55
     admiring
    0.54
    ړئ
    0.54
    除く
    0.54
     Роди
    0.54
    POSITIVE LOGITS
    berra
    0.70
    opies
    0.64
    oughby
    0.64
    amette
    0.63
    캔
    0.62
    hemian
    0.61
     कैन
    0.59
    adians
    0.59
     vary
    0.58
    terbury
    0.58
    Activations Density 0.195%

    No Known Activations