© 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. 38428
    Prev
    Next
    INDEX
    Explanations

    groups and their characteristics

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

    The neuron detects copular attribution statements—phrases where a subject (often a social group or “we”) is linked via “is/are/known to be” to a descriptive trait.

    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
     you
    -1.22
     your
    -1.09
     someone
    -0.99
     WARRANTIES
    -0.95
     time
    -0.95
     there
    -0.94
    ará
    -0.94
    你这个
    -0.91
     here
    -0.91
     new
    -0.90
    POSITIVE LOGITS
     their
    1.53
     they
    1.21
     themselves
    1.15
    他们
    1.10
    their
    1.09
    typically
    1.07
     généralement
    1.06
     mereka
    1.05
     geralmente
    1.05
    sfx
    1.04
    Activations Density 0.057%

    No Known Activations