© 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. 32913
    Prev
    Next
    INDEX
    Explanations

    Q:

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

    This neuron activates on legal boilerplate headings and phrases, especially “Terms of Use”/“Terms and Conditions” sections and similar standardized legal section markers.

    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
     will
    -1.16
     such
    -1.06
     hubiese
    -1.02
     with
    -0.98
    bufio
    -0.98
     desmotiv
    -0.93
    dersfield
    -0.93
     linguri
    -0.92
     at
    -0.92
     to
    -0.92
    POSITIVE LOGITS
    Here
    1.38
     you
    1.31
     tässä
    1.21
     hereinafter
    1.20
    “
    1.14
     financières
    1.12
     apprécié
    1.11
     YOU
    1.10
    dotto
    1.10
     présentes
    1.09
    Activations Density 0.083%

    No Known Activations