© 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. 128220
    Prev
    Next
    INDEX
    Explanations

    This neuron detects words signaling controversy or scandal, such as “controversial,” “damning,” and “embroiled.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    scandal and allegations

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    only
    -1.08
     also
    -1.04
    这也
    -1.02
    ׇ
    -1.00
     Pfar
    -1.00
     bluse
    -1.00
     giac
    -0.99
    Herzlich
    -0.97
     quae
    -0.95
     impostor
    -0.95
    POSITIVE LOGITS
     reputation
    1.47
     his
    1.40
     allegations
    1.24
     personagens
    1.16
     OGSÅ
    1.09
    เรื่อง
    1.08
     recent
    1.07
     wewnętr
    1.07
    疑惑
    1.07
     skle
    1.05
    Activations Density 0.094%

    No Known Activations