© 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. 40058
    Prev
    Next
    INDEX
    Explanations

    nationality/group followed by descriptor

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

    The neuron fires on meaningful content words—nouns, verbs, pronouns, and proper nouns—while ignoring common function words.

    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
    Such
    -1.12
    ereka
    -1.05
     such
    -1.00
    Výhody
    -0.93
     receive
    -0.91
    Both
    -0.88
     zostać
    -0.88
    fible
    -0.88
     będ
    -0.87
     espont
    -0.85
    POSITIVE LOGITS
     augusti
    0.96
     --------
    0.96
    itize
    0.91
    鼋
    0.91
     spes
    0.91
    imensional
    0.90
     یک
    0.90
     ένα
    0.90
    😃
    0.90
    uskas
    0.89
    Activations Density 0.036%

    No Known Activations