© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 66713
    Prev
    Next
    INDEX
    Explanations

    children learning and growing

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

    The neuron fires strongly on open‐class, content-bearing words (nouns, verbs, adjectives) and stays inactive on function words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     then
    -1.41
    וּ
    -1.17
    וּ
    -1.17
    breviations
    -1.14
    リコン
    -1.14
    消失了
    -1.11
    opardi
    -1.09
     should
    -1.09
     to
    -1.08
    ᴜ
    -1.07
    POSITIVE LOGITS
    ERSHIP
    1.42
     learning
    1.38
    ığımız
    1.34
    expandindo
    1.33
     daily
    1.32
    Inconvénients
    1.30
    ׇ
    1.28
     each
    1.27
     cabra
    1.27
     our
    1.24
    Activations Density 0.019%

    No Known Activations