© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 12796
    Prev
    Next
    INDEX
    Explanations

    ASCII diagrams and formatted text

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

    The neuron tends to pick out low-frequency, content-specific terms—especially technical nouns—in formal or patent-style text.

    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
    甚至
    0.88
    ce
    0.79
    etc
    0.73
    甚至是
    0.73
    м
    0.73
     etc
    0.71
    haps
    0.70
    觀
    0.70
    ِي
    0.69
    حدد
    0.68
    POSITIVE LOGITS
    ﮑ
    0.93
     ہوگا۔
    0.86
     ..$
    0.85
    HashSet
    0.84
    Ϲ
    0.81
     এছ
    0.79
     CASES
    0.78
    frameNStart
    0.77
     PERCENT
    0.77
     VESSELS
    0.77
    Activations Density 0.000%

    No Known Activations