© 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. 22-GEMMASCOPE-RES-131K
    4. 93207
    Prev
    Next
    INDEX
    Explanations

    as youas sampleas partas individualas itsas :

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

    This neuron specializes in spotting occurrences of the word “as” (e.g., in constructions like “published as,” “stored as,” “deposited as,” “upload as”).

    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
    がなく
    -1.45
     لإ
    -1.27
     passando
    -1.24
    见的
    -1.24
    があ
    -1.21
     durante
    -1.20
     to
    -1.19
    そのため
    -1.18
    偶尔
    -1.18
     dando
    -1.17
    POSITIVE LOGITS
     a
    2.97
     an
    2.09
     either
    1.58
     both
    1.50
     something
    1.41
     our
    1.39
     from
    1.34
     on
    1.27
     in
    1.27
     using
    1.26
    Activations Density 0.069%

    No Known Activations