© 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. 24-GEMMASCOPE-2-RES-16K
    7. 15000
    Prev
    Next
    INDEX
    Explanations

    The neuron responds to infrequent or out‐of‐vocabulary tokens—i.e. it lights up on rare, unusual, or unexpected words.

    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
    <h1>
    0.89
    من
    0.88
    連
    0.88
    day
    0.87
    def
    0.86
    core
    0.86
    empire
    0.82
    include
    0.80
    nect
    0.78
    <h2>
    0.77
    POSITIVE LOGITS
     может
    0.95
     отличается
    0.91
     TensorFlow
    0.86
    可以
    0.84
     Rayleigh
    0.82
     hace
    0.82
     echt
    0.82
    ர்ச்சி
    0.82
     выпол
    0.81
     Leonid
    0.81
    Activations Density 0.000%

    No Known Activations