© 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. 8816
    Prev
    Next
    INDEX
    Explanations

    so you, so I

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

    The neuron strongly reacts to the subordinating conjunction pair “so … that,” marking purpose or consequence clauses.

    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
    ك
    1.14
    ка
    0.87
    א
    0.85
    ع
    0.84
    us
    0.78
    شاء
    0.76
    क
    0.76
    িক
    0.75
    ুল্লাহ
    0.72
    我已经
    0.72
    POSITIVE LOGITS
    ledad
    1.17
     ώστε
    1.03
    특별시
    0.97
     authorizing
    0.90
    ્ટ
    0.88
     personalise
    0.85
    ار
    0.84
     pudi
    0.84
    रियो
    0.82
     reconnu
    0.82
    Activations Density 0.094%

    No Known Activations