© 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. 41-GEMMASCOPE-2-RES-262K
    7. 101853
    Prev
    Next
    INDEX
    Explanations

    reflex

    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_all/layer_41_width_262k_l0_big
    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
     carbides
    0.42
     Nge
    0.41
     носи
    0.41
     carbide
    0.39
     heavily
    0.38
    anjut
    0.38
    癡
    0.38
     Nd
    0.36
    ioctl
    0.35
     Bd
    0.35
    POSITIVE LOGITS
     reflex
    1.73
     reflexes
    1.59
     Reflex
    1.56
     reflexive
    1.45
    reflex
    1.44
     refleks
    1.43
     réflex
    1.38
    反射
    1.09
     rifless
    1.07
     refle
    0.90
    Activations Density 0.002%

    No Known Activations