© 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. 201354
    Prev
    Next
    INDEX
    Explanations

    metadata system

    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
     =
    0.46
     crashes
    0.45
     =\
    0.45
     demonstrates
    0.44
     verschiedene
    0.44
     evac
    0.44
     strikeouts
    0.43
     reinforces
    0.43
    ACO
    0.43
    নমেন্ট
    0.43
    POSITIVE LOGITS
     장
    0.49
    ophyll
    0.46
    osp
    0.44
    ༠
    0.44
    тите
    0.42
     시장
    0.42
     둘
    0.42
     Johnny
    0.41
     없고
    0.41
    ini
    0.40
    Activations Density 0.002%

    No Known Activations