© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 1070
    Prev
    Next
    INDEX
    Explanations

    safety and caution

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

    This neuron responds strongly to abstract virtue and religious/spiritual terms (e.g. peace, goodwill, mercy, faith, love).

    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_31_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
     हालांकि
    0.54
    各
    0.52
    ข้อมูล
    0.51
    頬
    0.50
    每
    0.50
     "
    0.49
     поскольку
    0.49
    以及
    0.49
    เนื่อง
    0.48
    萼
    0.48
    POSITIVE LOGITS
     permeates
    0.98
     prevails
    0.90
     ensues
    0.89
     transcends
    0.86
     isn
    0.86
     Happens
    0.85
     comes
    0.81
     thrives
    0.81
     Yourself
    0.80
     Comes
    0.79
    Activations Density 0.893%

    No Known Activations