© 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. 204248
    Prev
    Next
    INDEX
    Explanations

    expensive calculation, tension piles, combined analysis, supplements, visual pairing

    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
    тва
    0.50
    ми
    0.49
    Boxing
    0.47
     disastrous
    0.46
     visualization
    0.46
    핵
    0.45
    при
    0.45
     પ્રકાશ
    0.44
     conséquences
    0.44
    の変化
    0.44
    POSITIVE LOGITS
    rd
    0.53
    为了
    0.52
    ld
    0.50
     כדי
    0.46
    phi
    0.46
    เพื่อให้
    0.45
     Чтобы
    0.45
     该
    0.44
     disbursements
    0.44
     यहां
    0.43
    Activations Density 0.002%

    No Known Activations