© 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. 24-GEMMASCOPE-2-RES-16K
    7. 5759
    Prev
    Next
    INDEX
    Explanations

    concepts that are complex or difficult

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

    The neuron fires strongly on the first major content word of a sentence or clause—essentially singling out the sentence‐opening topic term.

    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
    သွား
    0.59
    glia
    0.59
     शूट
    0.57
    ڤ
    0.57
    丁寧に
    0.57
    έλ
    0.57
    μπ
    0.55
     ফোনে
    0.55
     प्रशंसा
    0.55
    റെ
    0.54
    POSITIVE LOGITS
     synonymous
    0.86
     suatu
    0.82
     synonyms
    0.82
     synonym
    0.77
     overrated
    0.73
     bedeutet
    0.71
     onderdeel
    0.71
    是一种
    0.68
     означает
    0.68
     a
    0.68
    Activations Density 0.407%

    No Known Activations