© 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. 22-GEMMASCOPE-RES-131K
    7. 27912
    Prev
    Next
    INDEX
    Explanations

    syntax error expected

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

    The neuron flags tokens that belong to program or parser error/exception messages (e.g. “Error,” “Wrong number of arguments,” “syntax error,” “invalid token,” “expected,” 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
    Vielleicht
    -1.01
     firft
    -0.99
    נחנו
    -0.99
    همان
    -0.97
     xvii
    -0.96
    pantalones
    -0.96
     powietrza
    -0.96
     detalló
    -0.96
     naše
    -0.95
     cilvē
    -0.94
    POSITIVE LOGITS
     chré
    1.03
     cannot
    0.96
    遑
    0.95
    usepackage
    0.94
    lış
    0.92
     at
    0.92
    皈
    0.92
    кон
    0.92
    व
    0.92
    зор
    0.89
    Activations Density 0.020%

    No Known Activations