© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 4372
    Prev
    Next
    INDEX
    Explanations

    This neuron detects statements of mandatory requirements or rules, flagged by words like “must,” “only,” or similar directive language.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Biraz
    0.73
     Mereka
    0.72
     Entire
    0.68
     Huge
    0.67
     那
    0.66
     Idee
    0.65
    やって
    0.63
     Cooperation
    0.63
     Ours
    0.63
    왔
    0.63
    POSITIVE LOGITS
    1
    0.86
     également
    0.83
    5
    0.82
    9
    0.81
    3
    0.79
    4
    0.77
    6
    0.76
     inoltre
    0.75
    7
    0.75
    2
    0.72
    Activations Density 0.111%

    No Known Activations