© 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. 66100
    Prev
    Next
    INDEX
    Explanations

    JSON parsing

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

    The neuron is picking out programming‐style tokens—variable and function names, keywords, and other code identifiers—rather than ordinary English words.

    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
     diversos
    -0.96
     fortement
    -0.94
     arbeitet
    -0.93
     diventa
    -0.89
     différents
    -0.89
     detalle
    -0.89
     beeld
    -0.88
     verschillende
    -0.87
     воспроиз
    -0.86
     ime
    -0.85
    POSITIVE LOGITS
     with
    1.33
     by
    1.05
     before
    1.02
     juta
    0.99
     it
    0.96
    toDouble
    0.96
    xxxxxx
    0.96
     actually
    0.95
     sœurs
    0.93
    💋
    0.92
    Activations Density 0.010%

    No Known Activations