© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 81546
    Prev
    Next
    INDEX
    Explanations

    prior art or appeal

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

    The neuron fires on isolated technical identifiers and other non-ordinary words—things like code tokens, file/URL components, library/type names, and similar “programming” or technical terms.

    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
    })_{
    -0.80
     GOP
    -0.79
     spectra
    -0.75
    atori
    -0.75
     kwa
    -0.75
     Dorf
    -0.69
     Darts
    -0.69
     Daarom
    -0.69
     logiciel
    -0.68
    gott
    -0.68
    POSITIVE LOGITS
     herausgegeben
    0.94
     desplazamiento
    0.78
     insertion
    0.76
     acoper
    0.75
     reemplazo
    0.74
     Beobachtungen
    0.73
     Entfernung
    0.73
     chóng
    0.71
    dibility
    0.69
     Ophthal
    0.68
    Activations Density 0.043%

    No Known Activations