© 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. 54420
    Prev
    Next
    INDEX
    Explanations

    specialized scientific or technical terms

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

    The neuron responds to prominent content words—especially specialized terms, proper nouns, and key numeric or technical identifiers—i.e. it highlights the main “topic” words in each sentence.

    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
     desafio
    -1.02
    ఝ
    -0.98
    头顶
    -0.94
     niedrig
    -0.90
     espesor
    -0.88
    のですね
    -0.87
    ч
    -0.86
    垆
    -0.86
    thode
    -0.85
     pruebas
    -0.85
    POSITIVE LOGITS
    \\
    0.88
     at
    0.85
     ידי
    0.83
     With
    0.82
    duplicate
    0.78
    лимпий
    0.77
    handleError
    0.77
    ;
    0.77
    rpt
    0.77
    нтен
    0.76
    Activations Density 0.040%

    No Known Activations