© 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. 48104
    Prev
    Next
    INDEX
    Explanations

    years, dates

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

    The neuron fires on numeric tokens—particularly numbers used in legal citations, section references, years, or footnote identifiers.

    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
    actually
    -1.70
    鵙
    -1.67
     ſever
    -1.61
     hongos
    -1.55
    ఽ
    -1.55
     fluide
    -1.47
    generally
    -1.47
     Dinas
    -1.45
    really
    -1.43
    wonderful
    -1.43
    POSITIVE LOGITS
     these
    2.22
     while
    1.96
     for
    1.90
     When
    1.82
     several
    1.80
     many
    1.77
     new
    1.77
     when
    1.75
     this
    1.70
     numerous
    1.70
    Activations Density 0.044%

    No Known Activations