© 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. 15721
    Prev
    Next
    INDEX
    Explanations

    questions following Q

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

    The neuron strongly activates on CJK (Chinese/Japanese/Korean) characters, i.e. non-Latin script 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_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
    chaise
    -1.27
     baratas
    -1.22
    TURE
    -1.14
    laient
    -1.13
     péro
    -1.12
    ikti
    -1.10
     Úl
    -1.09
    $,
    -1.09
     vín
    -1.08
    itern
    -1.08
    POSITIVE LOGITS
    f
    1.27
    ani
    1.19
    na
    1.11
    li
    1.11
    ns
    1.11
    ff
    1.10
    co
    1.09
    ot
    1.09
    c
    1.09
    id
    1.06
    Activations Density 0.000%

    No Known Activations