© 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. 10-GEMMASCOPE-RES-131K
    7. 39394
    Prev
    Next
    INDEX
    Explanations

    aws auto, ethics argue, cleanliness and

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

    The neuron is sensitive to named entities or titles—especially multi-word, capitalized names (e.g. league or stadium names).

    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
     is
    -1.82
     there
    -1.65
     There
    -1.51
     was
    -1.35
     it
    -1.32
    there
    -1.32
     without
    -1.29
    There
    -1.28
     these
    -1.27
     on
    -1.21
    POSITIVE LOGITS
    莛
    1.70
     pleaſure
    1.70
     Jerusalén
    1.62
     lenne
    1.55
     alemania
    1.51
     comprennent
    1.50
     compositeur
    1.50
     oude
    1.50
     nhưng
    1.48
    solete
    1.48
    Activations Density 0.111%

    No Known Activations