© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 8545
    Prev
    Next
    INDEX
    Explanations

    The neuron responds most strongly to rare or infrequent tokens—that is, unusual words or terms that hardly ever appear.

    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_24_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
    ून
    0.71
    #
    0.66
    BON
    0.66
     sogen
    0.65
     thậm
    0.64
     বন্দর
    0.63
     Medizin
    0.63
    МИ
    0.59
    وفر
    0.59
     échelle
    0.59
    POSITIVE LOGITS
    อ
    0.93
    อังกฤษ
    0.86
    կ
    0.86
    നും
    0.83
    ک
    0.81
    邓
    0.79
    ต
    0.78
    ვა
    0.75
    }^{[
    0.75
    պ
    0.75
    Activations Density 0.000%

    No Known Activations