© 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. 84525
    Prev
    Next
    INDEX
    Explanations

    xion, input, Conference, Hash, CLO, VEC, aged

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

    The neuron strongly activates on multi‐token proper names or titled phrases (e.g. person or place names, section headings, product/model names).

    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
    vermectin
    -1.44
    unno
    -1.36
    banez
    -1.35
    various
    -1.34
     geschenk
    -1.34
    licopter
    -1.29
    OTROS
    -1.28
     allerlei
    -1.27
    viembre
    -1.27
    zzie
    -1.26
    POSITIVE LOGITS
     frumos
    1.36
     one
    1.33
     as
    1.23
     можно
    1.23
     bobina
    1.20
     vzduchu
    1.20
    ,''
    1.17
     cesta
    1.16
     -
    1.14
     gmin
    1.13
    Activations Density 0.211%

    No Known Activations