© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 21261
    Prev
    Next
    INDEX
    Explanations

    people who

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

    The neuron strongly activates on the word “People” when it begins a phrase describing or categorizing a group (e.g. “People who…,” “People from…,” etc.).

    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
    The
    -2.38
    \
    -2.28
    遜
    -1.86
    大概
    -1.83
    不必
    -1.82
     son
    -1.82
    ͎
    -1.82
    кали
    -1.80
    痒
    -1.78
    posé
    -1.77
    POSITIVE LOGITS
    people
    2.70
    余談
    2.36
    鵟
    2.23
    joyed
    2.22
    ⸨
    2.20
    也将
    2.13
    葑
    2.11
    
    2.11
     広場
    2.06
    優れた
    2.00
    Activations Density 0.116%

    No Known Activations