© 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. 34-GEMMASCOPE-RES-131K
    7. 75489
    Prev
    Next
    INDEX
    Explanations

    discussing abstract concepts

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

    This neuron responds to common English function words—especially articles (a, the), conjunctions/relative pronouns (that, of), and question words (how, why)—rather than content 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_34/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
     fromage
    -1.00
    鑪
    -0.98
    OSIS
    -0.91
    èlement
    -0.88
     pochette
    -0.86
    ்
    -0.86
    chedelic
    -0.85
     tempel
    -0.85
     ścian
    -0.83
     hilfiger
    -0.82
    POSITIVE LOGITS
     football
    1.78
     baseball
    1.63
     esimerkiksi
    1.60
     soccer
    1.59
     pizza
    1.55
     basketball
    1.52
     golf
    1.48
     coffee
    1.42
     bijvoorbeeld
    1.41
     guitar
    1.41
    Activations Density 0.123%

    No Known Activations