© 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. 317
    Prev
    Next
    INDEX
    Explanations

    D &, W illing, S ATA

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

    The neuron strongly activates on uppercase words or acronyms (proper names, titles, and other all‐caps 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
     be
    0.27
    u
    0.26
    ar
    0.24
    en
    0.24
    ש
    0.24
    er
    0.24
    :
    0.23
    al
    0.23
    el
    0.22
    ak
    0.21
    POSITIVE LOGITS
    ﺎ
    0.30
    ﻮ
    0.23
    Ла
    0.23
    Λ
    0.21
    acterial
    0.21
    А
    0.21
    ১০
    0.20
    attering
    0.20
    Ü
    0.20
    umping
    0.19
    Activations Density 0.961%

    No Known Activations