© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14306
    Prev
    Next
    INDEX
    Explanations

    code-related keywords and structures

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

    This neuron lights up on code-like identifier tokens (e.g. class or function names, XML tags, import names) rather than ordinary natural-language words.

    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
     zwier
    -1.30
     цвят
    -1.17
    ֶּ
    -1.16
     intenção
    -1.13
     garagem
    -1.11
     mojej
    -1.09
     къде
    -1.08
    pernic
    -1.08
     privacidade
    -1.07
     llavero
    -1.06
    POSITIVE LOGITS
     or
    1.24
    旋律
    1.17
    便利
    1.10
     auch
    1.10
    "]=
    1.03
     vitales
    1.02
     bloque
    1.00
     yöntem
    1.00
     hanya
    1.00
     или
    1.00
    Activations Density 0.018%

    No Known Activations