Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-3-27B-IT
    3. 31-GEMMASCOPE-2-RES-262K
    4. 2499
    Prev
    Next
    INDEX
    Explanations

    uncompromising security

    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
     거고
    0.53
     people
    0.52
     because
    0.51
    と思いますが
    0.51
     примерно
    0.51
     যেহেতু
    0.50
    を使って
    0.50
    /
    0.50
    していますが
    0.48
     irgendwie
    0.48
    POSITIVE LOGITS
     столь
    0.61
     undeniably
    0.51
     bestowed
    0.49
     erneut
    0.49
     സൃഷ്ട
    0.48
    établ
    0.48
     unparalleled
    0.47
     uncompromising
    0.46
    Cette
    0.46
     unrival
    0.45
    Activations Density 0.118%

    No Known Activations