© 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. 22-GEMMASCOPE-RES-131K
    7. 15724
    Prev
    Next
    INDEX
    Explanations

    political power chip

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

    The neuron strongly activates on the opening headline and lead‐in sentences of news articles (especially political/world news).

    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
    poxy
    -1.27
     because
    -1.25
    ősz
    -1.24
    Eventually
    -1.22
    toluene
    -1.16
    propylene
    -1.16
     дів
    -1.14
    tijds
    -1.09
    dateien
    -1.09
    ecause
    -1.06
    POSITIVE LOGITS
     البته
    1.14
     expected
    1.09
     Mittag
    1.07
     hopes
    1.06
     baden
    1.04
     inside
    1.03
     strategy
    1.02
     nadzie
    1.02
     gewüns
    0.99
     Monday
    0.99
    Activations Density 0.061%

    No Known Activations