© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 36030
    Prev
    Next
    INDEX
    Explanations

    names and directions

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

    The neuron is primarily triggered by capitalized proper nouns—especially personal names.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    —
    -1.17
    --
    -1.06
     only
    -1.00
    僕の
    -0.94
     what
    -0.90
    -->
    -0.90
     frères
    -0.90
     that
    -0.90
     disheart
    -0.89
     дешевле
    -0.88
    POSITIVE LOGITS
    Zubereitung
    1.13
    Krank
    1.05
    nehin
    1.03
     perda
    1.01
     Kalender
    0.99
    җ
    0.99
    Демографія
    0.99
    ofollow
    0.97
    rative
    0.95
    шься
    0.95
    Activations Density 0.000%

    No Known Activations