© 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. 72612
    Prev
    Next
    INDEX
    Explanations

    spatial locations and directions

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

    The neuron specializes in spotting tokens that describe physical movement or travel through space (e.g. “crossed,” “through,” “driving to,” “walking through,” 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_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
     you
    -1.17
     this
    -1.07
     letzten
    -1.06
     also
    -1.06
     pastillas
    -0.97
    닙
    -0.93
     last
    -0.93
     suport
    -0.91
    wörter
    -0.91
     hydroxyl
    -0.90
    POSITIVE LOGITS
     street
    1.29
    hésite
    1.13
     conocía
    1.13
    &=&\
    1.10
     знако
    1.07
     известно
    1.06
    ítě
    1.05
    ± 
    1.05
     trató
    1.05
    iranje
    1.05
    Activations Density 0.010%

    No Known Activations