© 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. 1525
    Prev
    Next
    INDEX
    Explanations

    must be

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

    The neuron fires on occurrences of the modal “must” marking mandatory requirements or obligations in policy‐style text.

    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
    IENT
    0.51
    ANO
    0.50
    ﻨ
    0.49
    ajući
    0.49
    որ
    0.48
    ﺌ
    0.48
    անդ
    0.47
    <unused1728>
    0.47
    タミン
    0.47
    ﻤ
    0.47
    POSITIVE LOGITS
     belongs
    0.70
     merupakan
    0.70
     berada
    0.68
     bevat
    0.68
     είναι
    0.67
     ήταν
    0.67
     belong
    0.66
     terdiri
    0.66
     belonged
    0.65
     étant
    0.64
    Activations Density 0.001%

    No Known Activations