© 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. 104732
    Prev
    Next
    INDEX
    Explanations

    acronyms and specific terms

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

    The neuron fires strongly on standalone uppercase or title‐style tokens—things like section headings, acronyms, category labels, and prominent words in titles.

    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
     SDLK
    -1.03
    tabilité
    -0.97
    Dichlorobenzene
    -0.96
     estúdio
    -0.94
     SCHEME
    -0.94
     Produkten
    -0.93
     ondas
    -0.93
    maxAge
    -0.93
    iteiten
    -0.92
     Serviço
    -0.91
    POSITIVE LOGITS
    </strong>
    1.05
     nedenle
    0.99
    かず
    0.96
     ögon
    0.94
     then
    0.93
     ťaž
    0.90
     sonata
    0.88
     miljoner
    0.87
     människor
    0.86
     herbst
    0.86
    Activations Density 0.004%

    No Known Activations