© 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
    2. Gemma-3-27B-IT
    3. 22-GEMMASCOPE-2-TRANSCODER-262K
    4. 258394
    Prev
    Next
    INDEX
    Explanations

    The neuron detects phrases indicating quantity followed by a qualifier of variety. For instance, "over X different types" or "over Y species". The core pattern is establishing a large number or a broad category followed by a descriptor of diversity.Let's refine this based on the lists:- `MAX_ACTIVATING_TOKENS`: `over` is prominent. `0` is also frequent. `have`, `of`.- `TOKENS_AFTER_MAX_ACTIVATING_TOKEN`: `different`, `types`, `видов`. This strongly points to "different types" or "species".- `TOP_ACTIVATING_TEXTS`: Examples are "over 100 different types of HPV", "over 126 species of pine", "over 20,000 known species of bees", "nearly 40 different species of dolphin", "over 200 types of HPV", "over 10,000 known species of birds", "over 1,700 different mutations", "over 3,000 mutations", "over 7,500 varieties of apples".over X different types/species

    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_22_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
    Fixture
    0.45
     :
    0.41
    ería
    0.40
    णु
    0.40
     square
    0.39
     Crab
    0.38
    :
    0.38
     :
    0.37
     extortion
    0.36
     doubts
    0.36
    POSITIVE LOGITS
    赭
    0.46
    %';
    0.43
     Deleg
    0.42
     ડે
    0.42
     myelin
    0.41
    पालन
    0.41
    itsar
    0.40
     तेलंगाना
    0.39
     Sebagai
    0.39
    🅘
    0.39
    Activations Density 0.003%

    No Known Activations