© 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. 34-GEMMASCOPE-RES-131K
    7. 91743
    Prev
    Next
    INDEX
    Explanations

    GraphQL queries and variables

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

    The neuron strongly activates on code‐style tokens (identifiers, punctuation, strings, and paths) in programming snippets, effectively detecting source code segments.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     engines
    -0.81
    pling
    -0.76
    levels
    -0.71
     liberate
    -0.71
    kaar
    -0.70
     gonna
    -0.70
    pike
    -0.70
     cada
    -0.69
    ьера
    -0.69
     Diagnose
    -0.69
    POSITIVE LOGITS
     variables
    1.19
     gql
    1.14
     Variables
    1.13
    variables
    1.11
    Variables
    1.11
     query
    0.98
    fragment
    0.93
     fragment
    0.92
    query
    0.92
     VARIABLES
    0.90
    Activations Density 0.007%

    No Known Activations