© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 99198
    Prev
    Next
    INDEX
    Explanations

    States, countries, and projects

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

    The neuron strongly activates on formal administrative and bureaucratic terms—like “information,” “analysis,” “planning,” “decision,” “hold,” and “nominations”—that commonly appear in government or regulatory notices.

    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
    ⾦
    -1.63
     Uniwersyte
    -1.63
    DAFTAR
    -1.57
     hábiles
    -1.55
     boletín
    -1.55
    谥
    -1.54
    暧
    -1.50
     Suomessa
    -1.50
     calentamiento
    -1.48
    unikasi
    -1.45
    POSITIVE LOGITS
     it
    2.03
     you
    1.40
     like
    1.34
     again
    1.28
     değil
    1.25
     wasn
    1.22
     didn
    1.20
     プレミアム
    1.19
     doesn
    1.19
     сразу
    1.18
    Activations Density 0.154%

    No Known Activations