© 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. 15483
    Prev
    Next
    INDEX
    Explanations

    code structures and lists

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

    This neuron flags programming‐language tokens—e.g. markup tags, keywords, and identifiers—distinguishing code snippets from ordinary text.

    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
    })]
    -0.90
     They
    -0.86
    ்த்த
    -0.81
    而非
    -0.80
     Weisheit
    -0.77
     pierdas
    -0.75
    的模样
    -0.75
    Kilder
    -0.75
    のでしょう
    -0.73
    webservice
    -0.73
    POSITIVE LOGITS
     różne
    0.92
    hermes
    0.91
     endet
    0.88
    बस
    0.88
     jesteśmy
    0.87
     fourn
    0.87
    extAlignment
    0.86
    椤
    0.86
    erobic
    0.84
    becomes
    0.83
    Activations Density 0.016%

    No Known Activations