© 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. 8092
    Prev
    Next
    INDEX
    Explanations

    aluminum frame, dual learning, declarative web, methane concentrations, Effect Size, alpha channel, lucid sex

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

    This neuron fires on content-heavy topic words—i.e. key nouns that signal a subject or section theme.

    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
     niego
    0.67
    erweise
    0.66
    午後
    0.64
    ский
    0.63
     ۽
    0.63
    ®,
    0.60
    ские
    0.59
     quello
    0.59
    یک
    0.58
    が高い
    0.57
    POSITIVE LOGITS
     carcinogenesis
    0.69
     TABLE
    0.57
     शिवा
    0.56
     Mito
    0.55
     Sunil
    0.55
     千
    0.54
    OutputDir
    0.53
     JAS
    0.52
     Brod
    0.52
    ৃতি
    0.52
    Activations Density 0.208%

    No Known Activations