© 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. 22-GEMMASCOPE-RES-131K
    7. 25766
    Prev
    Next
    INDEX
    Explanations

    adjectives describe intensification or peculiarity

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

    The neuron strongly activates on very common English function words (e.g. “the,” “and,” “you,” “can,” “all,” “that,” “there,” etc.).

    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
     on
    -2.78
     in
    -2.73
     five
    -2.72
     or
    -2.72
     just
    -2.66
     two
    -2.66
     from
    -2.64
     when
    -2.59
     three
    -2.56
     eight
    -2.56
    POSITIVE LOGITS
     Dinas
    2.50
    despite
    2.22
     horrid
    2.20
     boister
    2.17
    particularly
    2.14
     soooo
    2.13
     ridiculously
    2.13
     goofy
    2.09
     insanely
    2.08
     wondrous
    2.06
    Activations Density 0.076%

    No Known Activations