© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 117617
    Prev
    Next
    INDEX
    Explanations

    word followed by descriptive word

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

    The neuron fires on semantically “heavy” or content‐bearing words—key nouns, verbs, and adjectives (e.g. proper names or topic-specific terms) rather than on 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_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
     maravill
    -1.29
     respetu
    -1.23
    さんで
    -1.18
     unbekannt
    -1.16
     sumamente
    -1.14
     sonriente
    -1.12
    getSelectedItem
    -1.12
     perfeito
    -1.11
     saliendo
    -1.11
     maravilh
    -1.09
    POSITIVE LOGITS
     manifesta
    1.20
     alcool
    1.20
     bateria
    1.16
     incess
    1.13
     mecan
    1.13
     trist
    1.13
     cabine
    1.11
     restaura
    1.11
    sprogram
    1.10
     conector
    1.09
    Activations Density 0.491%

    No Known Activations