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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 98542
    Prev
    Next
    INDEX
    Explanations

    safety and risk

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

    The neuron spotlights conditional warning or advisory phrases—i.e. “if …, do (not) …” style constructions.

    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
    之内
    -0.88
    ဟ
    -0.87
    только
    -0.87
    avoie
    -0.85
     voyons
    -0.85
    FAKE
    -0.84
    UML
    -0.83
     faudra
    -0.83
    其实
    -0.83
    hints
    -0.82
    POSITIVE LOGITS
     safety
    1.44
     risk
    1.39
     precautionary
    1.37
     safe
    1.30
     uncertainty
    1.28
     безопас
    1.23
    Risk
    1.23
     safer
    1.16
     precaution
    1.15
    risk
    1.13
    Activations Density 0.035%

    No Known Activations