© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 11874
    Prev
    Next
    INDEX
    Explanations

    This neuron detects strongly evaluative or judgmental words (e.g. surprise, impossible, trick, didn’t) that convey emphasis or certainty.

    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
     Crib
    0.46
     -&
    0.46
     కొంత
    0.46
     Thatcher
    0.46
     Özellikle
    0.45
     घेऊ
    0.44
    erdapat
    0.44
     thaliana
    0.44
    哪怕
    0.44
     कमजोर
    0.43
    POSITIVE LOGITS
    ！！
    0.66
    ？
    0.64
    ?",
    0.63
    ？！
    0.62
    ?";
    0.59
    ！？
    0.57
    !!!!
    0.57
    😂😂
    0.56
    "!
    0.56
    !!!"
    0.56
    Activations Density 0.215%

    No Known Activations