© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 1041
    Prev
    Next
    INDEX
    Explanations

    to + verb (infinitive)

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

    This neuron fires on the first salient content word that begins a new text segment or paragraph.

    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
     olish
    0.58
    ăt
    0.57
    ētu
    0.54
    ləşdir
    0.53
    itabbam
    0.52
    łego
    0.50
    <unused1748>
    0.50
     alleviate
    0.50
    쭐
    0.50
    <unused731>
    0.50
    POSITIVE LOGITS
     (
    0.63
    failed
    0.52
     pioneered
    0.50
     as
    0.49
     has
    0.49
    fought
    0.49
     fails
    0.48
     étant
    0.48
    ist
    0.48
     didn
    0.48
    Activations Density 0.000%

    No Known Activations