© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 12913
    Prev
    Next
    INDEX
    Explanations

    cat or categorizing

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron activates strongly whenever it sees the substring “cat” (e.g. as the standalone word “Cat” or as the prefix “cat” in a larger word).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    CRC
    0.55
    PATTERN
    0.53
    PSK
    0.52
    Noto
    0.52
    ‘
    0.52
    ポーツ
    0.51
    早就
    0.51
    蒻
    0.51
     কোথাও
    0.51
    Packard
    0.50
    POSITIVE LOGITS
     Cat
    1.07
     Kitty
    1.06
     cat
    1.04
     猫
    1.03
    猫
    1.01
    貓
    1.00
    cat
    0.99
     Cats
    0.97
    Kitty
    0.95
    Cat
    0.94
    Activations Density 0.201%

    No Known Activations