© 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. 40-GEMMASCOPE-2-RES-262K
    7. 192583
    Prev
    Next
    INDEX
    Explanations

    cautionary notes

    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_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
     '*':
    0.38
    hyde
    0.38
    median
    0.37
     '--
    0.37
     Trabal
    0.37
    禾
    0.37
     '${
    0.36
     prensa
    0.36
     crescent
    0.36
    무
    0.36
    POSITIVE LOGITS
    note
    0.60
    Note
    0.54
     note
    0.51
    NOTE
    0.50
    notes
    0.48
     Note
    0.47
    नोट
    0.47
     注
    0.45
     NOTE
    0.45
    Caution
    0.45
    Activations Density 0.007%

    No Known Activations