© 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. 10-GEMMASCOPE-RES-131K
    7. 11166
    Prev
    Next
    INDEX
    Explanations

    polynomials and proper nouns

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

    The neuron detects multi‐word or camel‐case technical names and proper terms (e.g. model/layer class names like “MobileNet” or “SeparableConv2D,” mathematical proper nouns such as “Temperley–Lie”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    sistence
    -1.31
    ⛥
    -1.27
     cuchillo
    -1.27
     choix
    -1.24
    瘣
    -1.23
     الألم
    -1.23
    鶺
    -1.23
     nivel
    -1.22
     notwithstanding
    -1.20
    ！】
    -1.20
    POSITIVE LOGITS
     besök
    1.47
    RESUMO
    1.41
     afhankelijk
    1.31
     çocuklar
    1.31
     hogyan
    1.30
     But
    1.28
     vanuit
    1.26
     utveckling
    1.26
     svå
    1.25
     förr
    1.25
    Activations Density 0.005%

    No Known Activations