© 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. 137131
    Prev
    Next
    INDEX
    Explanations

    understand low rates problems

    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
     foci
    0.43
    Ǧ
    0.41
    ？」
    0.38
     glandes
    0.38
    ܣ
    0.38
     मता
    0.37
    ける
    0.37
    櫃
    0.37
     penyeleng
    0.37
     trouble
    0.36
    POSITIVE LOGITS
     entendimento
    0.47
     understands
    0.46
    స్తుంది
    0.42
     ફક્ત
    0.41
    ará
    0.41
     voluntad
    0.41
    reibung
    0.40
    notin
    0.40
     khái
    0.40
    Understand
    0.40
    Activations Density 0.000%

    No Known Activations