© 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. 6048
    Prev
    Next
    INDEX
    Explanations

    words describing complexity

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

    The neuron is looking for the main “content” words in a sentence—i.e. the key verbs, nouns, and adjectives that carry its core descriptive or event meaning.

    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
    .
    0.74
     esas
    0.70
     
    0.69
    。
    0.68
     freaking
    0.68
     stb
    0.62
     extraordinaire
    0.61
    ..
    0.61
     similares
    0.60
     dle
    0.59
    POSITIVE LOGITS
    <unused2141>
    0.61
     असून
    0.59
    ၡ
    0.58
     metaphors
    0.57
    ቧ
    0.57
    <unused2148>
    0.57
    embangkan
    0.56
     isIn
    0.56
     phleg
    0.56
    <unused503>
    0.56
    Activations Density 0.651%

    No Known Activations