© 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. 10-GEMMASCOPE-RES-131K
    7. 12669
    Prev
    Next
    INDEX
    Explanations

    critically ill or stranded people

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

    The neuron specializes in spotting descriptive condition- or state-expressing words—especially adjectives or participial endings (e.g. “bald,” “ill,” “horny,” the “ing” and “ed” suffixes) that convey extreme or sensational states.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    ↵
    -2.06
     that
    -1.76
     the
    -1.66
    軚
    -1.44
     différentes
    -1.42
    </em>
    -1.39
     inoxid
    -1.38
     consideran
    -1.38
    u
    -1.36
    不由得
    -1.34
    POSITIVE LOGITS
     people
    2.39
     young
    2.02
     individuals
    1.98
     him
    1.84
     you
    1.78
     persons
    1.77
     Furcht
    1.75
     who
    1.67
     maravillosas
    1.67
     nineteen
    1.66
    Activations Density 0.318%

    No Known Activations