© 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. 13883
    Prev
    Next
    INDEX
    Explanations

    section, Protocol, Columns, my

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

    The neuron fires strongly on tokens characteristic of bibliography entries—things like commas, reference numbers, proper names, journal abbreviations, and numeric citations.

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

    punctuation marks, particularly commas, that appear between clauses or items in formal or technical writing.

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

    common function words and basic punctuation that serve as sentence glue rather than content.

    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
    LOAD
    0.86
     оборудования
    0.83
     противоре
    0.83
    Arquivo
    0.80
    লোড
    0.79
     склады
    0.79
     insbesondere
    0.77
     equipamento
    0.77
    ificação
    0.76
    arantad
    0.76
    POSITIVE LOGITS
     public
    0.77
     hard
    0.77
    ↵
    0.73
     creative
    0.72
     pesky
    0.72
     gun
    0.72
    िम
    0.71
     nal
    0.69
     marketing
    0.68
     nifty
    0.68
    Activations Density 0.001%

    No Known Activations