© 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. 4113
    Prev
    Next
    INDEX
    Explanations

    instructions, questions, directions

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

    The neuron fires on the curly apostrophe character (’), i.e. it flags tokens that are just the apostrophe in contractions or possessives.

    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
    ↵
    0.48
    The
    0.41
    the
    0.41
    One
    0.40
    Sugar
    0.39
    2
    0.39
    ↵↵
    0.38
    3
    0.38
    Sebastian
    0.38
    Ла
    0.37
    POSITIVE LOGITS
    ख्य
    0.43
    垩
    0.43
    परेशन
    0.40
     susp
    0.38
     commut
    0.38
     ჩ
    0.38
     ماين
    0.38
    emptive
    0.38
    ರ್ಭ
    0.38
    ismiss
    0.37
    Activations Density 0.000%

    No Known Activations