© 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. 93935
    Prev
    Next
    INDEX
    Explanations

    compound phrases after keyword

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

    The neuron selectively fires on multi‐syllable content words that are proper nouns or names of events, organizations, or locations.

    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
     contrário
    -0.86
    formKey
    -0.84
     osserva
    -0.81
     Hoof
    -0.80
     behaup
    -0.77
     intelig
    -0.75
    dawg
    -0.74
    くださ
    -0.73
    нению
    -0.73
    イエロー
    -0.72
    POSITIVE LOGITS
     awhile
    0.97
    rindo
    0.88
     Blanco
    0.85
     cum
    0.85
     cuarta
    0.84
     Bengaluru
    0.82
    imagined
    0.80
     kép
    0.79
     OSHA
    0.79
     Tout
    0.78
    Activations Density 0.045%

    No Known Activations