© 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. 10-GEMMASCOPE-RES-131K
    7. 58996
    Prev
    Next
    INDEX
    Explanations

    vertical bar |

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

    The neuron strongly detects the vertical bar character “|” (as used in mathematical notation, e.g. in set‐builder or absolute‐value expressions).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     Because
    -1.69
    жными
    -1.67
     Also
    -1.66
    ätigung
    -1.62
     so
    -1.60
     Obtén
    -1.55
    翔
    -1.55
     There
    -1.55
    。『
    -1.55
     because
    -1.53
    POSITIVE LOGITS
    2
    3.11
     you
    2.11
     of
    2.09
    あっという間に
    2.00
    1
    1.91
    真的好
    1.78
    あんまり
    1.74
    Jegy
    1.70
    Krank
    1.70
     amd
    1.68
    Activations Density 0.009%

    No Known Activations