© 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. 51373
    Prev
    Next
    INDEX
    Explanations

    vanishing gradient problem

    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
     H
    0.44
     Biblioteca
    0.44
     အ
    0.42
     Use
    0.42
     An
    0.41
     Amber
    0.41
     (+
    0.40
     လက်
    0.40
     وما
    0.40
     ,
    0.40
    POSITIVE LOGITS
     lawyers
    0.47
    ༦
    0.47
    টারে
    0.46
    acyjny
    0.45
     nějak
    0.45
    놨
    0.44
    ผู้ตัดสิน
    0.44
     meerdere
    0.43
     egregious
    0.42
    律师
    0.42
    Activations Density 0.021%

    No Known Activations