© 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. 34-GEMMASCOPE-RES-131K
    4. 119305
    Prev
    Next
    INDEX
    Explanations

    walking and movement

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

    The neuron detects mentions of athletic competition distances, especially numeric values paired with race-walking or race-walking distance units.

    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
    loten
    -0.77
     budowy
    -0.76
    stadion
    -0.73
     gradova
    -0.71
    vanta
    -0.70
    γωγ
    -0.69
     penyimpanan
    -0.68
    коменду
    -0.68
     bescha
    -0.68
    abr
    -0.68
    POSITIVE LOGITS
     walking
    7.16
     walk
    6.69
    walking
    6.06
    walk
    5.84
     Walking
    5.81
     walks
    5.78
    Walking
    5.72
     WALK
    5.63
     walked
    5.63
     Walk
    5.44
    Activations Density 0.119%

    No Known Activations