© 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. 71122
    Prev
    Next
    INDEX
    Explanations

    dirt, debris, harmful, virus

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

    words related to contaminants or harmful substances that need to be filtered, removed, or protected against.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron activates on words naming unwanted particulates or contaminants (e.g. dirt, dust, debris, germs, bacteria, soot).

    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
    Alguns
    -1.39
     видит
    -1.38
    ☆、
    -1.36
     Ultimately
    -1.35
     Shortly
    -1.34
      (
    -1.30
     departement
    -1.30
     Knowing
    -1.28
     it
    -1.27
    Horário
    -1.27
    POSITIVE LOGITS
     you
    1.41
     musisz
    1.41
     innym
    1.38
     godziny
    1.35
     pomaga
    1.34
     your
    1.34
     pozosta
    1.34
     jazdy
    1.34
     eerste
    1.33
     ใน
    1.33
    Activations Density 0.075%

    No Known Activations