© 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. 10957
    Prev
    Next
    INDEX
    Explanations

    position velocity

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

    This neuron responds to infrequent, content-heavy tokens (e.g. uncommon proper names, numbers, specialized terms) rather than common function words.

    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
     กี้
    0.88
    आ
    0.88
     ज
    0.86
    ج
    0.86
    ज
    0.83
    ה
    0.80
    గ
    0.78
    ک
    0.76
    😜
    0.76
    ل
    0.75
    POSITIVE LOGITS
     vyk
    0.70
     Tenants
    0.70
     Vek
    0.68
    ното
    0.68
     Hafen
    0.67
     Augustine
    0.65
     Accred
    0.65
     Vy
    0.65
    식회사
    0.65
     Oy
    0.63
    Activations Density 0.000%

    No Known Activations