© 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. 205119
    Prev
    Next
    INDEX
    Explanations

    initially chosen experience

    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
     Cas
    0.46
    엽
    0.45
    Ꮒ
    0.44
     Fl
    0.42
    них
    0.41
     substitutions
    0.41
     اللا
    0.41
    리
    0.40
    닥
    0.40
     Austral
    0.40
    POSITIVE LOGITS
    ributors
    0.47
     jendela
    0.46
     preparada
    0.46
    itorial
    0.46
     খেল
    0.46
    ma
    0.45
     tada
    0.45
    odată
    0.45
    ymmetrical
    0.45
    Mua
    0.45
    Activations Density 0.002%

    No Known Activations