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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-262K
    7. 6377
    Prev
    Next
    INDEX
    Explanations

    conflicting or mixed-language phrases

    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
    繽
    0.43
    槳
    0.40
    枨
    0.40
    媽媽
    0.39
    鲔
    0.38
    雩
    0.38
     வண்ணப்பூச்சு
    0.38
    鈮
    0.36
    赜
    0.36
    醌
    0.36
    POSITIVE LOGITS
    的事情
    0.37
     wronged
    0.36
     offend
    0.35
     offended
    0.35
     ****
    0.35
     revenge
    0.33
     avenge
    0.32
     humiliated
    0.32
     disguised
    0.31
     outsiders
    0.31
    Activations Density 0.003%

    No Known Activations