© 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. 385
    Prev
    Next
    INDEX
    Explanations

    This neuron fires strongly on the single-word or short-phrase headings of bulleted or enumerated action items (e.g. “Help,” “Activities,” “Offer,” “Appointments”) that introduce suggestions or instructions.

    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
     jonka
    0.28
     [](
    0.26
     whose
    0.25
     wherein
    0.25
    ().
    0.24
     INNER
    0.24
     ताकि
    0.24
     พร้อม
    0.24
    ።
    0.24
    whose
    0.24
    POSITIVE LOGITS
     είναι
    0.59
     is
    0.58
     isn
    0.53
     är
    0.50
     may
    0.50
     adalah
    0.50
     seems
    0.50
     tends
    0.48
    往往
    0.46
     není
    0.46
    Activations Density 9.112%

    No Known Activations