© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 2289
    Prev
    Next
    INDEX
    Explanations

    used to / once

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

    This neuron responds strongly to the phrase “used to,” marking expressions of past habitual actions or states.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ుల
    0.93
    ispiels
    0.92
     nėra
    0.91
    ශ්‍ය
    0.89
    )';
    0.89
    uana
    0.89
    ij
    0.88
    િંગ
    0.88
     változat
    0.88
    cheidet
    0.87
    POSITIVE LOGITS
     ago
    0.97
     stod
    0.95
     времена
    0.95
     americano
    0.93
     adage
    0.93
     άνθρω
    0.93
    le
    0.91
    이었다
    0.91
    땄
    0.89
     stabbing
    0.88
    Activations Density 0.015%

    No Known Activations