© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 127229
    Prev
    Next
    INDEX
    Explanations

    nouns describing specific things

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

    The neuron flags concrete entity nouns—especially institutions, roles, devices, or places (e.g. “police,” “server,” “company,” “town”).

    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
    
    -1.74
    非常に
    -1.73
    曏
    -1.69
     diantaranya
    -1.68
     Handeln
    -1.68
     podoba
    -1.67
    蘑
    -1.66
    ズニー
    -1.66
    Their
    -1.65
     musia
    -1.63
    POSITIVE LOGITS
     our
    1.60
     after
    1.57
     has
    1.55
    ↵↵
    1.54
     is
    1.45
     you
    1.38
     kedua
    1.33
     does
    1.31
     N
    1.30
    OC
    1.30
    Activations Density 0.104%

    No Known Activations