© 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. 24-GEMMASCOPE-2-RES-16K
    4. 6166
    Prev
    Next
    INDEX
    Explanations

    explanation code blocks

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

    The neuron detects high-level section headings or document structure labels (e.g. “Job Description,” “Summary,” “Responsibilities,” “Description”) in formal text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    section headings or labels in structured documents, particularly in job postings and technical documentation.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    formal section headings that label parts of structured documents (e.g., job postings, product descriptions, research abstracts).

    oai_token-act-pair · gpt-5Triggered 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
    ছাত্র
    0.66
     WORD
    0.63
    饗
    0.63
    Time
    0.62
    笊
    0.62
    ্ড
    0.61
    relevant
    0.60
     $-
    0.59
     $-$
    0.58
     DOI
    0.57
    POSITIVE LOGITS
    :“
    0.88
    :
    0.82
    这是一个
    0.82
     synopsis
    0.75
    Описание
    0.75
    :
    0.74
     Description
    0.73
    ک
    0.73
    Description
    0.73
     amacı
    0.72
    Activations Density 0.010%

    No Known Activations