© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 11945
    Prev
    Next
    INDEX
    Explanations

    extracting values or data

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

    The neuron is sensitive to code tokens naming data structures and manipulation methods (e.g. “data,” “json,” “map,” “values,” “DataFrame,” 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_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
     sguardo
    -0.75
     relates
    -0.75
    枕
    -0.74
     estratégica
    -0.74
     DATED
    -0.71
     determinada
    -0.70
     relacionada
    -0.70
     attendee
    -0.69
     estimado
    -0.69
     microbiota
    -0.69
    POSITIVE LOGITS
     values
    3.61
    values
    3.34
    Values
    2.75
     Values
    2.50
     VALUES
    2.34
     valeurs
    2.33
    VALUES
    2.25
     valores
    2.20
     wartości
    1.99
     Werte
    1.85
    Activations Density 0.006%

    No Known Activations