© 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. 8894
    Prev
    Next
    INDEX
    Explanations

    languages and coding elements

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

    The neuron is highly sensitive to special‐purpose tokens and boundary markers—things like the “@” in email addresses or the initial words and punctuation that signal a new question or emphatic statement.

    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
    सुद्धा
    0.82
    미
    0.78
     baseApiPath
    0.78
    ড়যন্ত্র
    0.77
    보면
    0.77
     सुद्धा
    0.73
    ண்கள்
    0.72
    UMIRE
    0.71
     राजांनी
    0.71
    ቢ
    0.70
    POSITIVE LOGITS
     Что
    0.79
     Tia
    0.79
     Setiap
    0.78
     प्रत्ये
    0.77
     данным
    0.76
    Ошибка
    0.75
     houve
    0.74
    igence
    0.74
     Здравствуйте
    0.73
    erreur
    0.71
    Activations Density 0.000%

    No Known Activations