© 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. 98444
    Prev
    Next
    INDEX
    Explanations

    Ab, At, E prefixes

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

    The neuron selectively activates on longer, morphologically complex content words—especially those with Latinate suffixes (e.g. “dominance,” “production,” “begging”).

    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
    iella
    -0.96
    ोंने
    -0.95
     valign
    -0.94
    ଂ
    -0.94
    olge
    -0.92
    viers
    -0.92
    enst
    -0.91
     Gewicht
    -0.91
    Referințe
    -0.91
     delantera
    -0.90
    POSITIVE LOGITS
     didn
    1.31
     wasn
    1.19
     doesn
    1.16
     compétition
    1.09
     medica
    1.08
     wouldn
    0.96
     امریک
    0.93
     uchun
    0.93
    並沒有
    0.91
     Europeo
    0.90
    Activations Density 0.000%

    No Known Activations