© 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-TRANSCODER-262K
    7. 79753
    Prev
    Next
    INDEX
    Explanations

    ensure, remember, pay attention

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_31_width_262k_l0_small_affine
    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
     prostu
    0.40
     없이
    0.38
     simplemente
    0.37
     सिंपली
    0.37
     ছাড়াই
    0.37
     puedo
    0.36
    $};
    0.36
    ⢸
    0.36
     semplicemente
    0.36
    という
    0.36
    POSITIVE LOGITS
     ensure
    0.62
     Ensure
    0.58
     jangan
    0.57
     beware
    0.54
    記得
    0.54
    要注意
    0.53
    必ず
    0.51
    Ensure
    0.50
    อย่า
    0.50
     đừng
    0.49
    Activations Density 0.076%

    No Known Activations