© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 31-GEMMASCOPE-2-RES-262K
    7. 109864
    Prev
    Next
    INDEX
    Explanations

    Stopping actions and providing reinforcements

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    צור
    0.50
    ਓ
    0.49
    祟
    0.47
    чному
    0.45
    ნის
    0.45
    事件
    0.44
    contenido
    0.44
    ی
    0.44
    Evento
    0.44
    кать
    0.44
    POSITIVE LOGITS
     max
    0.52
    ullo
    0.48
     Supreme
    0.47
     Mahar
    0.47
     Koh
    0.47
     understand
    0.46
     మూడు
    0.45
     strateg
    0.44
    oz
    0.44
     maxime
    0.44
    Activations Density 0.001%

    No Known Activations