© 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. 8948
    Prev
    Next
    INDEX
    Explanations

    asset beta

    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
    েই
    0.46
    --){
    0.46
     অঞ্চলে
    0.46
     Од
    0.45
    ext
    0.45
     Ре
    0.45
     读取
    0.44
    ол
    0.44
     从
    0.44
     కూడా
    0.44
    POSITIVE LOGITS
    stimulating
    0.49
    zení
    0.46
     volition
    0.46
    텀
    0.45
    stimulation
    0.45
     dinn
    0.45
    hybrid
    0.44
    calorie
    0.43
     morgen
    0.43
     kalor
    0.43
    Activations Density 0.003%

    No Known Activations