© 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. 4884
    Prev
    Next
    INDEX
    Explanations

    current

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

    The neuron flags the single most uncommon or topic‐specific content word (i.e. a rare non–function word) in each passage.

    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
    కి
    0.82
     enteral
    0.82
     concordance
    0.80
    朱
    0.79
     Infer
    0.77
     therapeut
    0.76
    РО
    0.76
     oligonucle
    0.76
    超
    0.75
    ס
    0.75
    POSITIVE LOGITS
    erie
    0.81
    affiliated
    0.79
    i
    0.77
    coated
    0.74
    irts
    0.74
    ્ષ
    0.73
    ano
    0.72
    I
    0.72
    penalty
    0.72
    iq
    0.70
    Activations Density 0.001%

    No Known Activations