© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 10931
    Prev
    Next
    INDEX
    Explanations

    technical explanations, code, programming

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

    The neuron fires strongly on the very first word or token of each new paragraph or section, effectively signaling the start of a new content block.

    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_31_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
    }$.
    1.00
    هایی
    0.78
     microwave
    0.74
    yaan
    0.73
    ní
    0.69
    },\\
    0.69
    জন্য
    0.68
    inin
    0.68
    ן
    0.68
    mn
    0.68
    POSITIVE LOGITS
    チュラル
    0.91
    取出
    0.82
    来る
    0.79
    潅
    0.79
    取り
    0.79
    >|</
    0.78
    ান
    0.77
     значения
    0.77
    Anyways
    0.76
    ע
    0.75
    Activations Density 0.003%

    No Known Activations