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. 40-GEMMASCOPE-2-RES-262K
    4. 3275
    Prev
    Next
    INDEX
    Explanations

    model responses

    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_40_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
    ・・
    0.30
    .•
    0.28
    Allows
    0.27
    わせ
    0.25
    どちら
    0.25
    ➖➖
    0.25
    ・・・・
    0.25
    。『
    0.25
    ‥
    0.25
    Acknowledgment
    0.24
    POSITIVE LOGITS
     absolutely
    0.32
     Absolutely
    0.32
     assolutamente
    0.31
     Sure
    0.29
     Release
    0.29
     Alright
    0.28
    ile
    0.27
    โม
    0.27
    an
    0.26
    ok
    0.26
    Activations Density 0.107%

    No Known Activations