© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 8861
    Prev
    Next
    INDEX
    Explanations

    never + verb

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

    The neuron strongly responds to the negation “never” (especially at sentence beginnings).

    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
    س
    1.91
    ات
    1.66
    om
    1.65
    ة
    1.63
    od
    1.41
    🔹
    1.40
    ون
    1.38
     источ
    1.35
    '.$
    1.33
    ség
    1.32
    POSITIVE LOGITS
     jamás
    1.70
    theless
    1.52
     stata
    1.48
     sekal
    1.44
     an
    1.43
     Never
    1.40
     NEVER
    1.36
    ਬ
    1.30
     Đến
    1.27
     a
    1.27
    Activations Density 0.072%

    No Known Activations