© 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. 10-GEMMASCOPE-RES-131K
    4. 19004
    Prev
    Next
    INDEX
    Explanations

    reported speech or past actions

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

    This neuron highlights mentions of actions or predictions—particularly verbs like wrote, learned, does, reads, expect—often referring to what the author or others have done or forecast.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     unserer
    -1.69
     felicitación
    -1.54
     dacă
    -1.54
    
    -1.52
     火鍋
    -1.50
    
    -1.48
     nếu
    -1.45
     奶茶
    -1.43
     gants
    -1.42
    グルト
    -1.41
    POSITIVE LOGITS
     those
    1.80
     NOT
    1.50
     ”
    1.38
     not
    1.31
    以上の
    1.30
     $
    1.27
     informações
    1.27
    ž
    1.25
     about
    1.24
     their
    1.23
    Activations Density 0.106%

    No Known Activations