© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-65K
    7. 1772
    Prev
    Next
    INDEX
    Explanations

    It's describing something

    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_40_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
     دیں۔
    0.22
     aident
    0.21
    ().
    0.20
    ”。
    0.20
    Yea
    0.20
     toning
    0.19
    "/
    0.19
    "。
    0.19
    ᆫ
    0.19
     ನಡೆ
    0.19
    POSITIVE LOGITS
     essentially
    0.33
     also
    0.30
     certainly
    0.29
     therefore
    0.26
     basically
    0.25
     represents
    0.25
     consists
    0.25
     serves
    0.25
     esche
    0.24
     functioned
    0.23
    Activations Density 0.566%

    No Known Activations