© 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. 10813
    Prev
    Next
    INDEX
    Explanations

    The neuron is essentially a “content‐word” detector: it ignores common function words (and punctuation) and lights up strongly on more semantically informative tokens (nouns, verbs, adjectives, etc.) rather than on little grammatical filler words.

    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
     comporte
    0.88
     verzek
    0.86
     innebär
    0.83
    NgramModel
    0.82
     belangrijkste
    0.77
     dezelfde
    0.75
    頁
    0.74
    ;")
    0.74
     wichtigen
    0.74
     combinatie
    0.73
    POSITIVE LOGITS
    y
    0.82
    ל
    0.75
    াড়া
    0.74
     সক্রিয়
    0.72
     เฟ
    0.71
    י
    0.71
    ɬ
    0.71
     Sada
    0.70
     Lear
    0.70
    SCREEN
    0.70
    Activations Density 0.000%

    No Known Activations