© 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. 62686
    Prev
    Next
    INDEX
    Explanations

    about positive choices

    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
     optimized
    0.47
     injunctive
    0.45
     optimised
    0.43
     Optimized
    0.43
    ignées
    0.41
    રૂ
    0.41
     ín
    0.41
    閩
    0.41
     ilç
    0.40
     African
    0.39
    POSITIVE LOGITS
     Sorry
    0.50
    Sorry
    0.50
    asty
    0.42
    saraba
    0.42
    aught
    0.41
    chris
    0.40
     sorry
    0.39
     そして
    0.37
    erals
    0.36
    pathetic
    0.36
    Activations Density 0.001%

    No Known Activations