© 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. 10470
    Prev
    Next
    INDEX
    Explanations

    insert before or after

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

    The neuron responds to rare or low‐frequency tokens—particularly proper names and uncommon multi‐syllable words.

    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
    al
    1.14
    L
    1.11
    H
    1.09
    ok
    1.07
    io
    1.00
     H
    0.98
    art
    0.97
    stadt
    0.96
    B
    0.96
     &
    0.92
    POSITIVE LOGITS
    י
    1.00
    Якщо
    0.94
     considérer
    0.89
    ць
    0.88
    ירים
    0.87
     URLs
    0.86
     інших
    0.86
    ך
    0.85
    יה
    0.85
     contrace
    0.82
    Activations Density 0.000%

    No Known Activations