© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 64604
    Prev
    Next
    INDEX
    Explanations

    thinking about

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

    The neuron flags semantically-rich content words (nouns, verbs, adjectives, adverbs) and ignores common function 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_22/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
     успешно
    -1.06
    不愿意
    -0.97
    materi
    -0.96
     تاکید
    -0.96
    successfully
    -0.93
     cœurs
    -0.91
     vœux
    -0.90
    ㍑
    -0.89
    ライター
    -0.88
     insiste
    -0.88
    POSITIVE LOGITS
     about
    2.00
     think
    1.43
    ^{
    1.16
    Think
    1.12
     Think
    1.12
     things
    1.02
     wobec
    1.00
     fondly
    0.99
    ldots
    0.97
     acerca
    0.96
    Activations Density 0.061%

    No Known Activations