© 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. 6282
    Prev
    Next
    INDEX
    Explanations

    date and time formatting

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

    The neuron is most sensitive to descriptive modifier words—especially prominent adjectives and adverbs that shift or qualify the discourse.

    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
    '
    0.83
    :</
    0.82
    {(-
    0.79
    ০
    0.78
    </
    0.76
    jsonplaceholder
    0.76
    '।
    0.75
    〔
    0.75
    এখন
    0.74
    aż
    0.74
    POSITIVE LOGITS
    𝘺
    0.89
     heng
    0.81
     arrears
    0.76
     stamp
    0.75
     pries
    0.75
     Weddings
    0.75
    𝒹
    0.74
    त्सा
    0.73
     bero
    0.73
    balik
    0.73
    Activations Density 0.000%

    No Known Activations