© 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. 33726
    Prev
    Next
    INDEX
    Explanations

    left cross

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

    The neuron is triggered by directional step instructions (e.g. “step left,” “step right”) in dance- or movement-style sequences.

    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
     all
    -0.83
     generally
    -0.83
     even
    -0.83
    part
    -0.82
     first
    -0.82
     arra
    -0.79
    imuth
    -0.79
     seemingly
    -0.78
     at
    -0.77
     those
    -0.77
    POSITIVE LOGITS
     ALEX
    0.97
     dernières
    0.96
     ayur
    0.90
     теперь
    0.88
     ᾿
    0.86
     jeżeli
    0.85
    ządz
    0.84
    FIB
    0.82
     Achieve
    0.82
    irii
    0.82
    Activations Density 0.014%

    No Known Activations