© 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-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 73744
    Prev
    Next
    INDEX
    Explanations

    words before languages

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

    The neuron is selectively activating on content‐bearing (non–stopword) terms—i.e. substantive words that carry the core meaning of a passage.

    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
    УДК
    -1.39
     anaknya
    -1.38
     剑
    -1.38
    Тел
    -1.38
     döner
    -1.35
    rierung
    -1.35
     comté
    -1.33
    ֿ
    -1.33
    ГЛА
    -1.33
     设计
    -1.31
    POSITIVE LOGITS
     our
    2.08
     nuestra
    1.63
     regeringen
    1.48
     or
    1.47
     Konzert
    1.41
     dañ
    1.38
     
    1.38
    upuncture
    1.38
     pojawia
    1.36
     nosso
    1.34
    Activations Density 0.080%

    No Known Activations