© 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. 42421
    Prev
    Next
    INDEX
    Explanations

    plays as a or competes

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

    The neuron activates on verbs that describe a person’s professional activity or role (e.g., works, plays, competes).

    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
     such
    -1.33
     our
    -1.33
     cilvē
    -1.26
    については
    -1.25
     químico
    -1.24
     cuando
    -1.24
     something
    -1.22
     lorsque
    -1.22
     svolta
    -1.20
     each
    -1.17
    POSITIVE LOGITS
     maquillage
    1.32
     apartament
    1.25
     parfum
    1.21
    until
    1.14
     of
    1.13
    као
    1.12
     agricole
    1.10
     filma
    1.09
    ROSE
    1.09
    continence
    1.09
    Activations Density 0.240%

    No Known Activations