© 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. 48108
    Prev
    Next
    INDEX
    Explanations

    "upper" followed by location or medical term

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

    The neuron detects geographical location names, particularly with "Upper" followed by a place name (like "Upper East Side" or "Upper Lusatian").

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron consistently activates on occurrences of the word “upper” (in any context or capitalization).

    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
    kembangan
    -0.81
     aksi
    -0.81
    ทุก
    -0.79
    ज्ञान
    -0.78
    octopus
    -0.77
    chalten
    -0.76
     ara
    -0.75
    awaited
    -0.75
    不喜欢
    -0.74
    enerated
    -0.71
    POSITIVE LOGITS
    Upper
    1.19
     Upper
    1.17
    Bound
    1.13
    bound
    1.07
    most
    1.06
     bound
    1.05
    deck
    0.96
    upper
    0.96
     respiratory
    0.93
     tier
    0.93
    Activations Density 0.027%

    No Known Activations