© 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. 31-GEMMASCOPE-2-RES-16K
    7. 8499
    Prev
    Next
    INDEX
    Explanations

    Homeland Security, terrorism

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

    The neuron activates on mentions of formal government agencies, programs, and bureaucratic entities (e.g. DHS, DOJ, Byrne/JAG, grant or preparedness programs).

    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
    烧
    0.61
    修理
    0.57
     Accountants
    0.56
     mediating
    0.55
     réparation
    0.55
     ورو
    0.54
     tristeza
    0.54
     afl
    0.52
     repairing
    0.52
    鼓
    0.52
    POSITIVE LOGITS
     Homeland
    1.66
     homeland
    1.61
     terrorist
    1.48
     Terrorism
    1.48
     terrorism
    1.46
    Terror
    1.44
    terrorism
    1.41
     терро
    1.39
     terrorists
    1.34
     Terror
    1.33
    Activations Density 0.228%

    No Known Activations