© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 56-GEMMASCOPE-2-TRANSCODER-262K
    7. 24959
    Prev
    Next
    INDEX
    Explanations

    for

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_56_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     c
    0.85
     ص
    0.60
    speaking
    0.54
     speaking
    0.51
     Speaking
    0.49
    Speaking
    0.46
     ц
    0.42
    c
    0.40
     tracing
    0.40
    Considering
    0.39
    POSITIVE LOGITS
     focus
    0.43
     confining
    0.38
     chilling
    0.37
     note
    0.37
    focus
    0.37
     fokus
    0.37
     Zoom
    0.36
     sinking
    0.36
     Focus
    0.36
    ø
    0.36
    Activations Density 0.007%

    No Known Activations