© 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. 65868
    Prev
    Next
    INDEX
    Explanations

    code structure and function calls

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

    This neuron selectively activates on specialized technical noun tokens—especially long, domain‐specific terms common in scientific or technical writing.

    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
     niebieski
    -1.04
     biały
    -0.91
     tričko
    -0.90
    ğiniz
    -0.90
     beslut
    -0.89
    ější
    -0.88
     publicidade
    -0.87
     SEÑ
    -0.86
     anledning
    -0.86
     utafiti
    -0.86
    POSITIVE LOGITS
    檜
    0.88
    aknya
    0.84
    cija
    0.84
     its
    0.84
     Când
    0.79
    hova
    0.78
    भाव
    0.77
    just
    0.77
    /
    0.77
    学员
    0.75
    Activations Density 0.021%

    No Known Activations