© 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-65K
    7. 46311
    Prev
    Next
    INDEX
    Explanations

    offering definitely help

    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_65k_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
     very
    0.75
    s
    0.74
     unrival
    0.73
    К
    0.71
     delicious
    0.71
    G
    0.70
    すべての
    0.68
     unparalleled
    0.66
    क
    0.65
    ]
    0.65
    POSITIVE LOGITS
    is
    0.74
    谠
    0.56
    ený
    0.54
    ená
    0.54
    ాయ
    0.53
     phận
    0.53
    ové
    0.52
    ర్య
    0.51
    thenReturn
    0.50
    ié
    0.50
    Activations Density 0.039%

    No Known Activations