© 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. 34-GEMMASCOPE-RES-131K
    7. 65255
    Prev
    Next
    INDEX
    Explanations

    chemical compound names and properties

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

    technical or legal terms in scientific or legal documents.

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

    This neuron fires most strongly on formal section headers, citation markers, and other document-structure labels (e.g. “v.” in case names, “models” in paper titles, “License” headings).

    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
    реди
    -0.89
    🚌
    -0.83
    ǣ
    -0.82
     évé
    -0.82
    吞噬
    -0.82
     hiver
    -0.80
    ᵥ
    -0.79
    Созда
    -0.78
     Boiler
    -0.77
    Mans
    -0.77
    POSITIVE LOGITS
     being
    0.88
     actually
    0.87
    tii
    0.86
     Hinsicht
    0.81
     smile
    0.77
    NewReader
    0.75
     ostat
    0.74
     finally
    0.72
     apparently
    0.72
     tramite
    0.71
    Activations Density 0.003%

    No Known Activations