© 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. 11241
    Prev
    Next
    INDEX
    Explanations

    "Barty", "Jax", "Bea", "Croco", "Barnacle Bart" nicknames

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

    The neuron strongly activates on rare or content-heavy tokens—uncommon words, proper names, numeric data, and other distinctive “information-rich” tokens.

    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
    гнозы
    0.82
     khususnya
    0.81
    tura
    0.80
    âtres
    0.79
    larni
    0.79
    )。
    0.79
    iapan
    0.79
    )</
    0.78
    )}$.
    0.76
    lanır
    0.76
    POSITIVE LOGITS
    ’
    0.80
     Certification
    0.79
    ᅮ
    0.76
    ED
    0.75
    ast
    0.73
    AM
    0.73
    OT
    0.73
    EM
    0.71
     certification
    0.71
    ᆯ
    0.70
    Activations Density 0.001%

    No Known Activations