© 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. 10-GEMMASCOPE-RES-131K
    7. 83664
    Prev
    Next
    INDEX
    Explanations

    want to show

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

    The neuron fires on common English “glue” words—high-frequency function words like pronouns (it, he), auxiliaries (was, have), conjunctions (because, then), articles, negations, and similar grammatical tokens.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     couldn
    -1.43
     have
    -1.42
     instead
    -1.41
     will
    -1.39
     )
    -1.37
     or
    -1.36
     i
    -1.30
    ',
    -1.30
    ソナ
    -1.29
     s
    -1.28
    POSITIVE LOGITS
     bricola
    1.48
    eeee
    1.45
     ultime
    1.45
    eeeeeeee
    1.44
    eeeeee
    1.41
    iiiii
    1.38
    eeeee
    1.34
     ideale
    1.33
    athons
    1.30
     estadounid
    1.30
    Activations Density 0.077%

    No Known Activations