© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 22522
    Prev
    Next
    INDEX
    Explanations

    ballroom, gala, prom, ball events

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

    The neuron fires on mentions of formal dance events (e.g., "prom," "ball," "gala").

    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
     ſeveral
    -1.70
     SIX
    -1.55
    cant
    -1.55
     Thirteen
    -1.52
    includes
    -1.49
     TWENTY
    -1.48
    夘
    -1.46
    contains
    -1.45
     SEVEN
    -1.45
     EIGHT
    -1.43
    POSITIVE LOGITS
    𖤣
    1.42
     paja
    1.39
    Nights
    1.29
     doctrina
    1.25
    lgari
    1.23
    Sniper
    1.23
     пръ
    1.23
    Shrimp
    1.23
     You
    1.22
     problema
    1.21
    Activations Density 0.015%

    No Known Activations