© 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. 107660
    Prev
    Next
    INDEX
    Explanations

    hypothesize, conjecture, imagine, plan

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

    The neuron fires on mid‐word scientific or scholarly root morphemes (e.g. “hypothe-,” “conject-,” “plani-,” “rum-”) that appear in academic‐style vocabulary.

    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
     laine
    -0.92
    otrophic
    -0.87
     Vesti
    -0.86
    César
    -0.86
    Makan
    -0.84
     alumnos
    -0.83
     decide
    -0.82
     stát
    -0.81
     saha
    -0.81
    جابة
    -0.80
    POSITIVE LOGITS
    w
    1.02
    mag
    0.97
    M
    0.93
    Hard
    0.93
    Below
    0.93
    Mag
    0.91
    Sal
    0.91
    m
    0.89
    ว่า
    0.88
    hard
    0.86
    Activations Density 0.022%

    No Known Activations