© 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-RES-262K
    7. 55725
    Prev
    Next
    INDEX
    Explanations

    than comparable

    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
     sickening
    0.46
     oversees
    0.44
     regionally
    0.43
    㭚
    0.41
    赳
    0.40
     Chowdh
    0.40
    ത്തും
    0.39
     عكس
    0.39
     gute
    0.39
    𝖺
    0.39
    POSITIVE LOGITS
    A
    0.50
    Bea
    0.46
    parametric
    0.46
    D
    0.44
    Than
    0.43
    F
    0.43
    ,
    0.42
    FFFF
    0.42
    Cut
    0.41
    Panic
    0.41
    Activations Density 0.001%

    No Known Activations