© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 47458
    Prev
    Next
    INDEX
    Explanations

    familial relationships involving in-laws

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

    This neuron responds to the hyphen character, i.e. tokens that connect words or form compound/negative constructions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     perlahan
    -2.03
     what
    -1.95
    there
    -1.80
    镌
    -1.65
     a
    -1.59
     any
    -1.58
    {
    -1.54
     increíble
    -1.52
     sengaja
    -1.52
     renang
    -1.50
    POSITIVE LOGITS
     
    2.05
    も同じ
    1.84
     \
    1.66
     focuses
    1.60
     cuja
    1.59
     [
    1.55
     exhilarating
    1.51
     EVEN
    1.48
     relentless
    1.48
     impor
    1.45
    Activations Density 0.002%

    No Known Activations