© 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. 68844
    Prev
    Next
    INDEX
    Explanations

    proper names and places

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

    This neuron responds to the repeated appearance of the page’s main proper-name heading (i.e. the article’s title) when it reoccurs as standalone tokens.

    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
     recently
    -1.09
    <0xF4>
    -1.09
    只有
    -1.08
    而
    -1.02
    那
    -1.02
    最近
    -1.01
    近
    -1.00
    medlem
    -0.98
    都
    -0.95
     izvē
    -0.93
    POSITIVE LOGITS
     and
    1.30
    oczes
    1.20
    Citiți
    1.01
    Enllaces
    0.96
     pierws
    0.95
     понравилось
    0.94
     Wikisource
    0.94
    ってみた
    0.93
    おっ
    0.93
    placa
    0.93
    Activations Density 0.045%

    No Known Activations