© 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. 24-GEMMASCOPE-2-RES-16K
    4. 4924
    Prev
    Next
    INDEX
    Explanations

    Implementing, visit, numbers, and conjunctions

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

    The neuron fires on duty‐statement or action‐item verbs (e.g. “Participate,” “Conduct,” “Prepare,” “Review”) that introduce tasks and responsibilities.

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

    This neuron specializes in spotting imperative action verbs that introduce instructions or procedural steps.

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

    This neuron detects the main action verbs (imperative or infinitive task statements) that begin duty or responsibility bullet points in job‐description–style text.

    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
    こういう
    0.82
     এরকম
    0.75
     একটা
    0.71
     biraz
    0.71
    뭐
    0.70
     trochu
    0.68
     얘
    0.68
    让我们
    0.68
    보다
    0.66
     이런
    0.65
    POSITIVE LOGITS
     and
    0.79
     establishes
    0.77
     மற்றும்
    0.73
     كافة
    0.71
     throughout
    0.71
     جميع
    0.69
    部長
    0.68
     maintains
    0.67
    جميع
    0.66
     during
    0.65
    Activations Density 0.290%

    No Known Activations