© 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. 16544
    Prev
    Next
    INDEX
    Explanations

    legal citations and case names

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

    The neuron responds strongly to unusually long runs of whitespace or blank‐space tokens (i.e. extended spaces/empty lines).

    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_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     just
    -1.43
     when
    -1.38
     first
    -1.32
     for
    -1.27
     let
    -1.26
     still
    -1.18
     only
    -1.17
     more
    -1.16
     that
    -1.15
     well
    -1.13
    POSITIVE LOGITS
     eenig
    1.05
    PostInfinity
    1.05
     intrigu
    1.05
    髂
    1.04
    utenant
    1.04
    בת
    1.03
     dát
    1.02
    itempty
    1.02
     trám
    1.02
    髁
    0.99
    Activations Density 0.000%

    No Known Activations