© 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. 8543
    Prev
    Next
    INDEX
    Explanations

    What does "

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

    The neuron fires strongly on unusually rare or idiosyncratic subword tokens (low‐frequency character fragments).

    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
    ры
    0.88
     Возможно
    0.84
    overlapping
    0.83
    тным
    0.80
    。『
    0.79
    льным
    0.79
     ممکن
    0.79
    ﺰ
    0.75
    рых
    0.75
    новен
    0.75
    POSITIVE LOGITS
     scritta
    0.78
     chama
    0.75
     chiam
    0.75
     chari
    0.75
     своїх
    0.75
     coaches
    0.73
     chast
    0.73
     অভাবে
    0.73
     кін
    0.71
     seine
    0.70
    Activations Density 0.000%

    No Known Activations