© 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. 6986
    Prev
    Next
    INDEX
    Explanations

    un- prefix for negative meaning

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

    The neuron flags modal and advisory language—i.e. words expressing necessity, possibility, or instruction (must, should, can, will, etc.).

    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
    涓
    0.59
     naï
    0.57
     επιχει
    0.56
     ingenu
    0.54
    自信
    0.54
     preempt
    0.54
    表面
    0.54
    Petit
    0.54
    浅
    0.53
    ልቅ
    0.53
    POSITIVE LOGITS
     disliked
    1.94
     bad
    1.93
     kötü
    1.92
     unpleasant
    1.88
     disagreeable
    1.88
    悪い
    1.87
     খারাপ
    1.85
     undesirable
    1.84
     xấu
    1.82
    糟糕
    1.82
    Activations Density 1.037%

    No Known Activations