© 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. 77006
    Prev
    Next
    INDEX
    Explanations

    section and introduction headers

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

    This neuron selectively activates on standalone numeric tokens (e.g., years, page or section numbers, counts).

    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
     exempel
    -0.95
    书籍
    -0.90
    ('*',
    -0.89
     –
    -0.88
    º
    -0.87
     ۱۸
    -0.86
     didn
    -0.84
    OutOfRange
    -0.82
    ­
    -0.82
     isn
    -0.81
    POSITIVE LOGITS
    Abstract
    1.93
     Abstract
    1.71
    ABSTRACT
    1.66
     abstract
    1.61
     ABSTRACT
    1.55
    doi
    1.41
     this
    1.41
    abstract
    1.36
     authors
    1.22
     Keywords
    1.18
    Activations Density 0.360%

    No Known Activations