© 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. 98270
    Prev
    Next
    INDEX
    Explanations

    specific noun followed by related noun

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

    The neuron fires strongly on the main “topic” word or subject of a sentence or post—typically a key noun or proper‐noun (e.g. My, Ghost, Buying, Mule, Grails).

    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
    ίη
    -1.09
     nerds
    -0.98
    ַּ
    -0.98
    brille
    -0.95
    tiert
    -0.95
    よろしくお願い
    -0.94
     flori
    -0.93
    Rez
    -0.93
    走吧
    -0.93
     and
    -0.93
    POSITIVE LOGITS
     religieuses
    1.21
    soldier
    0.98
    emble
    0.98
     Spielen
    0.94
     dianteiro
    0.91
    Ethylbenzene
    0.90
    imagination
    0.89
     encantado
    0.87
    いただける
    0.87
    spotlight
    0.86
    Activations Density 0.250%

    No Known Activations