© 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. 22-GEMMASCOPE-RES-131K
    7. 12946
    Prev
    Next
    INDEX
    Explanations

    titles and authors

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

    This neuron specializes in spotting metadata identifiers—particularly numeric sequences and section headings such as titles, patent or document numbers, and citation codes.

    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
     even
    -1.27
     for
    -1.17
     as
    -1.11
     can
    -1.09
     probably
    -1.08
     in
    -1.02
     despite
    -1.02
     nearly
    -1.02
     on
    -1.01
     would
    -1.01
    POSITIVE LOGITS
     てき
    1.16
    nowu
    1.15
     personagens
    1.14
    ilions
    1.13
     présentant
    1.13
     erbjuder
    1.13
     ouvertes
    1.13
     PÅ
    1.12
     anunció
    1.11
     trám
    1.11
    Activations Density 0.020%

    No Known Activations