© 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. 10-GEMMASCOPE-RES-131K
    7. 90441
    Prev
    Next
    INDEX
    Explanations

    good cop and the bad cop

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

    The neuron fires on contentful “open‐class” words (i.e. non–stop words: nouns, verbs, adjectives, etc.) rather than on common function words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     of
    -5.81
    ของ
    -1.50
    of
    -1.46
    weds
    -1.39
     about
    -1.38
    也会
    -1.36
     even
    -1.34
    UPAC
    -1.32
    мка
    -1.32
     it
    -1.30
    POSITIVE LOGITS
    CancelButton
    1.47
     llevan
    1.45
     destinada
    1.42
     funcionando
    1.39
     constituye
    1.38
    ╩
    1.38
     desean
    1.36
    嫵
    1.35
    uminense
    1.35
     same
    1.35
    Activations Density 0.133%

    No Known Activations