© 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
    5. 31-GEMMASCOPE-2-RES-262K
    7. 65435
    Prev
    Next
    INDEX
    Explanations

    softly rounded, as Samsung, is part

    np_acts-logits-general · gemini-2.5-flash-lite

    defining difficulty or purity

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-pt/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    is
    0.53
    us
    0.50
    н
    0.49
     =
    0.47
    ite
    0.46
    aw
    0.46
    ufthansa
    0.46
    setMaximum
    0.46
    max
    0.46
    у
    0.45
    POSITIVE LOGITS
     sería
    0.57
     quantidades
    0.55
     voisines
    0.54
     inconn
    0.53
     pediu
    0.53
     douloureuse
    0.53
     atque
    0.52
     seraient
    0.52
     soyez
    0.52
     pourront
    0.52
    Activations Density 0.000%

    No Known Activations