© 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. 10-GEMMASCOPE-RES-131K
    7. 105969
    Prev
    Next
    INDEX
    Explanations

    in or through

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

    The neuron is essentially detecting short directional/function words—especially the preposition “in” (and by extension similar adverbial particles like “up” or “through”).

    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
    戕
    -2.02
     sélectionner
    -1.95
    UARIO
    -1.92
    蟶
    -1.88
     všechny
    -1.86
     preparada
    -1.86
    どうして
    -1.84
     personnaliser
    -1.84
    纡
    -1.84
    わたしは
    -1.83
    POSITIVE LOGITS
     of
    2.78
     a
    2.50
     are
    2.38
     the
    2.30
     It
    2.27
     about
    1.91
     in
    1.88
    in
    1.88
     they
    1.86
     Either
    1.85
    Activations Density 0.001%

    No Known Activations