© 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. 31-GEMMASCOPE-2-TRANSCODER-262K
    7. 54735
    Prev
    Next
    INDEX
    Explanations

    The `MAX_ACTIVATING_TOKENS` list contains words like "Amazing", "incredible", "unusual", and "Extraordinary".The `TOP_POSITIVE_LOGITS` list also contains related terms like "невероят" (Russian for "incredible") and "extraordin".The `TOP_ACTIVATING_TEXTS` show examples like "The Amazing World of Gumball", "Remarkably Bright Creatures", "bestselling author, most notably of *The Code of the Extraordinary Mind*", and phrases describing sensational words/keywords like "shocking", "amazing", "incredible", "unbelievable", "mind-blowing".It also includes Russian phrases like "Придумай 10 самых необычных рецептов" (Invent 10 most unusual recipes) and English phrases using "fun and unusual request".The neuron seems to activate on words denoting exceptionality, surprise, or outstanding qualities, often appearing in contexts of lists, titles, or descriptions of exciting/unusual things.Let's refine:- "Amazing", "incredible", "Extraordinary", "unusual".exceptional or unusual qualities

    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_31_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
    িল্লি
    0.40
     bag
    0.37
     sectarian
    0.37
    bag
    0.36
     sog
    0.36
    +}$
    0.36
    anjang
    0.36
     bagging
    0.35
     incub
    0.34
    Bag
    0.34
    POSITIVE LOGITS
     CENT
    0.38
     невероят
    0.36
     오전
    0.35
    ING
    0.34
     Gases
    0.34
    CENT
    0.33
    вую
    0.32
    ადი
    0.32
     extraordin
    0.32
    ît
    0.32
    Activations Density 0.001%

    No Known Activations