© 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. 108166
    Prev
    Next
    INDEX
    Explanations

    Malta and malt-related words

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

    The neuron is highly sensitive to geographic place names—especially “Malta” and related island- or location-based terms.

    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
    cycle
    -0.89
     Parisi
    -0.88
     السلام
    -0.81
    ATV
    -0.77
     Ulrich
    -0.76
    لاة
    -0.75
    Côte
    -0.75
    cycles
    -0.75
     Đà
    -0.75
    мора
    -0.73
    POSITIVE LOGITS
     Malta
    1.93
     Maltese
    1.73
    Malta
    1.41
     malta
    1.19
     Маль
    1.05
     Malte
    0.97
     Malt
    0.94
     маль
    0.90
     Knights
    0.86
     Sovereign
    0.84
    Activations Density 0.028%

    No Known Activations