© 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. 33221
    Prev
    Next
    INDEX
    Explanations

    path beginnings like /usr/ or C:\

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

    This neuron fires on tokens that form filesystem paths or PATH‐style environment‐variable assignments (slashes, colons, tildes, equals signs, “PATH,” 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
    <0xE6>
    -1.09
     jätte
    -1.06
    jaguar
    -1.03
     legjobb
    -1.02
    Ẩ
    -1.01
     totalmente
    -1.00
    CHOOL
    -0.98
    aviation
    -0.98
     ibland
    -0.98
     maart
    -0.97
    POSITIVE LOGITS
     '-';
    1.01
     Pläne
    1.00
     Erkenntnis
    0.99
    ’;
    0.99
     shou
    0.97
     traf
    0.97
     odes
    0.94
    éléments
    0.93
    ttää
    0.91
     renegade
    0.90
    Activations Density 0.011%

    No Known Activations