© 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. 57509
    Prev
    Next
    INDEX
    Explanations

    past actions by people or systems

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

    The neuron detects mentions of people having gone through or experienced something (e.g. “experienced,” “been through,” “has done,” “anyone who…”).

    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
     There
    -1.81
     In
    -1.69
    There
    -1.68
     on
    -1.63
    7
    -1.57
     The
    -1.45
    In
    -1.43
     as
    -1.42
     had
    -1.41
     in
    -1.38
    POSITIVE LOGITS
    ambiance
    1.68
     ſeveral
    1.66
    Helps
    1.59
     Améli
    1.56
     clô
    1.52
     procha
    1.51
     Eſ
    1.51
     bụ
    1.50
    produces
    1.49
    mbgg
    1.48
    Activations Density 0.109%

    No Known Activations