© 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. 66666
    Prev
    Next
    INDEX
    Explanations

    parts of body or clothing

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

    This neuron is attuned to common grammatical “function” words—short, high-frequency items like articles (a, the), pronouns (his, my), prepositions (of, at), and similar stop-word tokens.

    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
    ઃ
    -1.01
    からです
    -0.97
     понадоби
    -0.92
     in
    -0.92
     because
    -0.91
    хой
    -0.90
    感じる
    -0.89
     allowable
    -0.88
     related
    -0.88
     from
    -0.86
    POSITIVE LOGITS
     τῆς
    1.15
     tipis
    1.15
     guantes
    1.10
    სქოლიო
    1.09
    jenkins
    1.07
    pleasant
    1.03
     assise
    1.00
     playera
    0.99
     Taxes
    0.98
     Gorg
    0.98
    Activations Density 0.007%

    No Known Activations