© 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. 34-GEMMASCOPE-RES-131K
    7. 57462
    Prev
    Next
    INDEX
    Explanations

    performing actions and movements

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

    The neuron fires on tokens that describe someone or something carrying out an action (e.g. “does,” “perform,” “working,” “sporting,” “performing,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ık
    -0.77
    lotten
    -0.77
    vis
    -0.77
    typal
    -0.76
     Babel
    -0.74
    cée
    -0.74
     альтерна
    -0.72
    ber
    -0.72
    unikasi
    -0.71
    det
    -0.70
    POSITIVE LOGITS
     spins
    1.58
     flips
    1.52
     pir
    1.52
     gymnastics
    1.38
     cart
    1.36
    🤸
    1.35
     spin
    1.33
     spinning
    1.32
     acrobat
    1.32
     performing
    1.25
    Activations Density 0.031%

    No Known Activations