© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 85837
    Prev
    Next
    INDEX
    Explanations

    religious and existential themes

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

    The neuron strongly activates on function words that serve as conjunctions or clause-linkers (e.g. “for,” “but,” “which,” “so”).

    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
    <bos>
    -1.94
    S
    -1.78
    L
    -1.69
     which
    -1.68
    P
    -1.64
    W
    -1.51
    M
    -1.42
    J
    -1.42
     nuovi
    -1.41
    哪里
    -1.39
    POSITIVE LOGITS
     décès
    1.50
     escalier
    1.47
    Если
    1.41
    Cette
    1.40
    На
    1.40
     cimetière
    1.37
    Вы
    1.36
    ’
    1.36
    Мы
    1.35
    Quelques
    1.35
    Activations Density 0.009%

    No Known Activations