© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 15774
    Prev
    Next
    INDEX
    Explanations

    explanations or contexts

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

    This neuron fires strongly on tokens within enthusiastic or promotional praise—e.g. words in exclamatory, complimentary, or “I would love/try” style positive-recommendation contexts.

    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
    idated
    0.74
     Этот
    0.71
     MyRegisterClass
    0.70
     Kantor
    0.69
     Kunststoff
    0.68
    篾
    0.68
     Questo
    0.68
    ärg
    0.67
     Corrosion
    0.67
     Reprodu
    0.66
    POSITIVE LOGITS
     cruel
    0.81
    ف
    0.80
    ل
    0.80
    ઝ
    0.76
     obligations
    0.74
    گ
    0.74
     einfach
    0.73
    ن
    0.73
     mindless
    0.71
     headwinds
    0.71
    Activations Density 0.110%

    No Known Activations