© 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. 95351
    Prev
    Next
    INDEX
    Explanations

    ordinal numbers followed by superlatives

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

    The neuron strongly activates on standalone identifier tokens—single-word labels or field names (e.g. “Engraving,” “Embossing,” “London,” “content,” “auxiliary”) used as metadata keys or tags in structured code or markup.

    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
     untuk
    -1.27
    已经
    -1.23
     sekarang
    -1.16
     technically
    -1.13
    已經
    -1.11
     üblichen
    -1.11
     empieza
    -1.10
    trab
    -1.10
    ADI
    -1.08
    想要
    -1.07
    POSITIVE LOGITS
    而非
    1.41
    两人的
    1.40
    他們的
    1.38
     teint
    1.37
    吅
    1.34
    谪
    1.32
     réessayer
    1.30
    龶
    1.30
     blancs
    1.28
    ("")
    1.26
    Activations Density 0.009%

    No Known Activations