© 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. 25550
    Prev
    Next
    INDEX
    Explanations

    specific issues or points

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

    The neuron fires on abstract plural nouns that typically introduce or enumerate list items (e.g. “issues,” “errors,” “points,” “changes,” “ideas”).

    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
    When
    -1.09
    Another
    -1.08
    Whether
    -1.05
    If
    -1.04
     who
    -1.02
     kovo
    -1.01
    Because
    -0.99
    ♪♪
    -0.96
    Two
    -0.95
    Дина
    -0.92
    POSITIVE LOGITS
     including
    1.13
     checklist
    1.11
     attributes
    1.11
     cualidades
    1.02
    bodied
    1.00
     simplesmente
    1.00
     aspectos
    1.00
     mejoras
    0.98
     conductas
    0.97
     listed
    0.96
    Activations Density 0.087%

    No Known Activations