© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 16008
    Prev
    Next
    INDEX
    Explanations

    syntactic elements

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

    The neuron strongly activates on content words that refer to entities—particularly personal pronouns (“you,” “I”) and named (often capitalized or technical) nouns.

    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
    ił
    0.88
    pared
    0.84
    雱
    0.84
    commended
    0.83
    angering
    0.83
     conspiring
    0.82
     deducting
    0.81
    एसएल
    0.80
    Jed
    0.79
     obstructing
    0.79
    POSITIVE LOGITS
    рия
    0.70
     ryt
    0.67
    нибудь
    0.63
     rarity
    0.63
     III
    0.62
    和小
    0.62
     和
    0.59
    мин
    0.59
     fucking
    0.58
     Waves
    0.58
    Activations Density 0.001%

    No Known Activations