© 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. 105895
    Prev
    Next
    INDEX
    Explanations

    method thereof

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

    The neuron fires most strongly on uncommon or domain-specific tokens—that is, relatively rare words.

    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
    ,"
    -1.72
    0
    -1.41
    ,
    -1.35
    <em>
    -1.32
    cessed
    -1.29
    єм
    -1.27
    =
    -1.26
     a
    -1.25
    all
    -1.23
     remains
    -1.23
    POSITIVE LOGITS
     operazioni
    1.53
    もありました
    1.51
    も多い
    1.50
    senza
    1.48
     dovol
    1.47
     voluto
    1.46
    Empereur
    1.46
    葙
    1.45
     potrzebu
    1.45
    entière
    1.44
    Activations Density 0.128%

    No Known Activations