© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 38829
    Prev
    Next
    INDEX
    Explanations

    required or mandatory actions

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

    The neuron fires on words and word‐fragments related to controlling, restricting, or regulating content and items (e.g. censor, self-regulate, screen, remove, warning).

    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
     décider
    -0.86
    宀
    -0.86
     rely
    -0.84
     secours
    -0.84
     joindre
    -0.82
     rembour
    -0.80
     lauter
    -0.79
     réserver
    -0.79
     обраща
    -0.78
     utilisez
    -0.76
    POSITIVE LOGITS
     required
    2.00
     Required
    1.80
    required
    1.75
     REQUIRED
    1.70
     mandatory
    1.70
    Required
    1.63
     demanded
    1.59
     forced
    1.56
    REQUIRED
    1.55
     compulsory
    1.48
    Activations Density 0.099%

    No Known Activations