© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 99721
    Prev
    Next
    INDEX
    Explanations

    action related to walking

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

    The neuron strongly activates on occurrences of the verb “walk” (and its inflected form “walked”)—especially when it forms part of a common phrasal expression (e.g. “walk in,” “walk out,” “walk on”).

    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
     running
    -1.01
    climbing
    -0.94
     climbing
    -0.93
     traveling
    -0.89
     riding
    -0.88
     travelling
    -0.86
     driving
    -0.86
    观点
    -0.85
     chạy
    -0.84
     Zwischen
    -0.82
    POSITIVE LOGITS
    🚶
    1.38
     Walks
    1.13
    걷
    1.03
     streets
    1.02
     paths
    1.02
    Walk
    0.99
     aisle
    0.96
     Walk
    0.96
     walked
    0.93
     corridors
    0.91
    Activations Density 0.023%

    No Known Activations