© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 53938
    Prev
    Next
    INDEX
    Explanations

    philosophical concepts, arguments, or conditions

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

    The neuron strongly activates on scholarly ethical or policy language expressing deontic concepts (e.g. “harm threshold,” “reproductive decision,” “organ donation,” “morally obligatory,” “duty,” “should”).

    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
     of
    -1.20
     malta
    -1.09
    TÜ
    -1.02
    ֪
    -1.02
    validation
    -0.99
    ђа
    -0.98
    ZnO
    -0.98
    нин
    -0.98
     оса
    -0.97
    subsubsection
    -0.97
    POSITIVE LOGITS
     lluvias
    1.20
    に行って
    1.20
     rocas
    1.19
     Unterricht
    1.18
     causar
    1.15
     dominar
    1.14
     derra
    1.13
     liburan
    1.13
     curar
    1.13
     maquin
    1.10
    Activations Density 0.058%

    No Known Activations