© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 9519
    Prev
    Next
    INDEX
    Explanations

    punctuation and line breaks

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

    The neuron is highly sensitive to backslash characters or escape‐sequence tokens (e.g. “\”, “runtime\”, “...\”) indicating code or file‐path–style formatting.

    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_31_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
    -',
    0.60
     persisted
    0.60
     అధికారు
    0.57
    Է
    0.54
    handlungen
    0.53
    렘
    0.51
     Dirección
    0.50
     değişt
    0.50
     डेब
    0.50
     estabele
    0.50
    POSITIVE LOGITS
    auteur
    0.55
    nons
    0.55
    UTTON
    0.54
     ಭಾಗ
    0.52
    走路
    0.52
    俑
    0.51
    愚
    0.51
    GROW
    0.49
    하여
    0.49
    驥
    0.49
    Activations Density 0.003%

    No Known Activations