© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 31-GEMMASCOPE-2-RES-262K
    7. 26676
    Prev
    Next
    INDEX
    Explanations

    improve performance, countless creatures, loss

    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
    +,
    0.54
     настолько
    0.50
     ONLY
    0.49
    且
    0.49
    、
    0.48
     with
    0.48
    していますが
    0.47
     as
    0.46
     এতটাই
    0.46
     ক্ষেত্রেই
    0.46
    POSITIVE LOGITS
    ܀
    0.63
     ។
    0.62
    。
    0.61
    ۔
    0.61
    ።
    0.59
    。」
    0.56
    》。
    0.55
    ។
    0.55
     हराया
    0.50
     ہے۔
    0.50
    Activations Density 0.182%

    No Known Activations