© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 10349
    Prev
    Next
    INDEX
    Explanations

    will or would

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

    The neuron strongly activates on mentions of formal legal proceedings (e.g. “tribunal,” “investigate,” “report to the President”), so it’s detecting references to judicial or investigatory bodies and actions.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     jsem
    0.72
     Myself
    0.68
     eigenlijk
    0.67
    なのですが
    0.67
     मैंने
    0.67
     navigated
    0.66
     నాకు
    0.64
     Workbook
    0.64
     Roughly
    0.64
    Ꮂ
    0.64
    POSITIVE LOGITS
     خواهند
    0.93
    就會
    0.92
    就会
    0.91
    will
    0.87
     되겠죠
    0.87
     будут
    0.83
     zullen
    0.82
     potranno
    0.82
    Will
    0.81
    きっと
    0.81
    Activations Density 0.049%

    No Known Activations