© 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-TRANSCODER-262K
    7. 27012
    Prev
    Next
    INDEX
    Explanations

    hold / believe views

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_40_width_262k_l0_small_affine
    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
     यूज
    0.40
     Connected
    0.37
     MPM
    0.37
     responding
    0.36
     paran
    0.36
     application
    0.36
    ガス
    0.36
    \{-\
    0.36
     نگر
    0.36
     तुलना
    0.35
    POSITIVE LOGITS
     hold
    0.62
     holds
    0.57
    hold
    0.52
    Holds
    0.51
     holding
    0.50
     Hold
    0.48
    holds
    0.47
     Holds
    0.47
    holding
    0.47
    Hold
    0.46
    Activations Density 0.014%

    No Known Activations