© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 8911
    Prev
    Next
    INDEX
    Explanations

    foreign language words

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

    The neuron fires on tokens carrying common suffix patterns—most notably place‐name endings like “-bourne” and web‐domain endings like “.com.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    ViewChild
    -1.72
    veno
    -1.71
     your
    -1.71
    的眼神
    -1.71
    this
    -1.65
    RIOS
    -1.64
    endforeach
    -1.63
     ſol
    -1.61
    
    -1.60
    your
    -1.59
    POSITIVE LOGITS
    Garantie
    1.58
    Stap
    1.56
    sandalia
    1.52
     слишком
    1.49
     сахара
    1.48
    所以在
    1.47
    Bibliografie
    1.45
    Geld
    1.45
    Abon
    1.45
    并未
    1.45
    Activations Density 0.003%

    No Known Activations