© 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. 22-GEMMASCOPE-RES-131K
    7. 67130
    Prev
    Next
    INDEX
    Explanations

    statistical research and patients

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

    the beginning of a new section or paragraph in academic or technical documents.

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

    This neuron fires on prominent section- or sentence-initial content words (like “In,” “Case,” “Analysis,” “Data,” “Effect,” etc.) that mark new sections or transitions in scientific or academic texts.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     competen
    -1.00
     kablo
    -0.95
     domina
    -0.94
     portait
    -0.93
     molle
    -0.92
     ilust
    -0.90
     dona
    -0.90
     biodivers
    -0.88
     recenti
    -0.88
     gire
    -0.87
    POSITIVE LOGITS
     it
    1.09
     shouldn
    1.04
     because
    1.00
     anyway
    0.91
     often
    0.91
     rowspan
    0.89
     cannot
    0.88
     especially
    0.86
     imagine
    0.84
     otherwise
    0.83
    Activations Density 0.091%

    No Known Activations