© 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. 10-GEMMASCOPE-RES-131K
    7. 21829
    Prev
    Next
    INDEX
    Explanations

    describing an action or process

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

    The neuron detects introductory “In …-ing” clauses—i.e. the pattern “In” followed by a gerund verb form.

    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
     in
    -2.45
     by
    -2.31
     when
    -1.88
     if
    -1.87
     before
    -1.83
     without
    -1.73
     had
    -1.71
     information
    -1.70
     went
    -1.57
     tiny
    -1.56
    POSITIVE LOGITS
     ſont
    1.88
    particularly
    1.82
    specifically
    1.80
     continually
    1.71
    
    1.60
    媗
    1.54
     perpé
    1.51
     ſou
    1.49
     soooo
    1.48
    艄
    1.47
    Activations Density 0.030%

    No Known Activations