© 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. 11651
    Prev
    Next
    INDEX
    Explanations

    the real benefit or power

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

    The neuron fires on the key action words and intensifying adverbs that open up advice or instruction sentences (e.g. “actually,” “reap,” “want,” “make”), effectively marking the start of a directive or motivational statement.

    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
     klikken
    0.55
    ্তন
    0.54
    跡
    0.52
    recon
    0.52
    okines
    0.51
     hồi
    0.51
     คะ
    0.50
    BackPressed
    0.50
     obnov
    0.50
    clicked
    0.49
    POSITIVE LOGITS
     Solaris
    0.62
     Wap
    0.59
     প্রতিষ্ঠ
    0.55
     Bis
    0.55
     Telf
    0.55
     Suzanne
    0.55
     TLS
    0.55
     Q
    0.54
     পৌরসভার
    0.54
     turbo
    0.54
    Activations Density 0.000%

    No Known Activations