© 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. 9809
    Prev
    Next
    INDEX
    Explanations

    not anything / nothing

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

    This neuron fires strongly on key words used in defining or promotional statements—verbs like “expect” and “means” that introduce mission, purpose, or core claims in marketing-style copy.

    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
    undaki
    0.60
    ကာ
    0.59
    どの
    0.55
    جميع
    0.54
    uraea
    0.51
    NR
    0.50
     جميع
    0.50
     (~
    0.49
    完全に
    0.49
    কো
    0.47
    POSITIVE LOGITS
     nothing
    3.98
    nothing
    3.73
     Nothing
    3.66
    Nothing
    3.66
     NOTHING
    3.61
     ничего
    3.20
     nada
    3.05
     nimic
    2.92
     nichts
    2.77
     rien
    2.69
    Activations Density 0.683%

    No Known Activations