© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 8507
    Prev
    Next
    INDEX
    Explanations

    introduces specific context/situation

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

    The neuron is tuned to spot numeric and quantitative expressions (e.g. numbers, percentages, measurements) in the text.

    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
     because
    -1.55
     when
    -1.54
     then
    -1.44
     who
    -1.36
     livelli
    -1.30
    ৄ
    -1.30
     what
    -1.28
     to
    -1.21
     According
    -1.19
     hưng
    -1.18
    POSITIVE LOGITS
     "
    1.52
     ."
    1.37
     vues
    1.36
    :"
    1.34
     :"
    1.33
    ю
    1.32
    蹑
    1.29
     "&
    1.28
     "/
    1.27
     ">
    1.27
    Activations Density 0.061%

    No Known Activations