© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 65933
    Prev
    Next
    INDEX
    Explanations

    film and music credits

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

    The neuron activates on words naming people’s roles or credits—e.g. occupational nouns (driver, trainer, owner) and production‐credit verbs (written, composed, produced, directed, arranged).

    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
     our
    -1.52
     useful
    -1.40
     volverá
    -1.29
    各有
    -1.28
     protecting
    -1.20
     they
    -1.20
     adequate
    -1.19
     people
    -1.18
     quickly
    -1.18
     decirle
    -1.17
    POSITIVE LOGITS
     diriger
    1.46
     brune
    1.45
     monstre
    1.44
     vider
    1.41
     oeil
    1.41
     astu
    1.34
     poë
    1.34
    ····
    1.31
     rente
    1.30
     pittores
    1.29
    Activations Density 0.058%

    No Known Activations