© 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. 127671
    Prev
    Next
    INDEX
    Explanations

    which + article or noun

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

    The neuron fires on occurrences of the relative pronoun “which,” especially when it begins a prepositional phrase (e.g. “under which,” “through which”).

    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
    );
    -2.42
     amusing
    -2.17
     intricate
    -2.11
    話の
    -2.06
    𓋼
    -2.03
     getting
    -1.98
     two
    -1.98
     first
    -1.97
     highlighted
    -1.93
     meticulous
    -1.92
    POSITIVE LOGITS
     they
    2.31
    烀
    2.06
     you
    2.06
     Herbs
    2.05
     it
    2.05
     オイル
    1.98
     possono
    1.92
    ğiniz
    1.91
     lakini
    1.89
    }.
    1.85
    Activations Density 0.035%

    No Known Activations