© 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. 20365
    Prev
    Next
    INDEX
    Explanations

    press Enter or try again

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

    This neuron activates on the main action verbs (the key “doing” words) in both narrative and instructional text.

    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
     would
    -1.63
     you
    -1.59
     should
    -1.50
     before
    -1.43
     when
    -1.43
     will
    -1.40
     during
    -1.35
     could
    -1.34
     might
    -1.30
     although
    -1.28
    POSITIVE LOGITS
     apparence
    1.29
     principe
    1.25
     bougies
    1.17
    初めての
    1.16
     harmonie
    1.16
     jammer
    1.13
    אח
    1.12
     chaises
    1.10
     kame
    1.10
    その他の
    1.10
    Activations Density 0.158%

    No Known Activations