© 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. 34-GEMMASCOPE-RES-131K
    4. 103010
    Prev
    Next
    INDEX
    Explanations

    legal and code context markers

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

    The neuron strongly activates on non-natural-language tokens—numbers, punctuation, symbols, code keywords, and other technical or structural markers—rather than on ordinary 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_34/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
    bited
    -0.74
    uchten
    -0.70
    Tib
    -0.69
    CCB
    -0.69
     Domains
    -0.69
    ţiei
    -0.68
    DataTo
    -0.67
    wasan
    -0.66
     زيادة
    -0.66
     Lois
    -0.66
    POSITIVE LOGITS
     natürlichen
    0.75
     STAT
    0.68
     réf
    0.67
    fizierung
    0.67
     tradizionale
    0.66
     referee
    0.64
    etary
    0.64
     Beginning
    0.64
    igas
    0.63
     inning
    0.63
    Activations Density 0.074%

    No Known Activations