© 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. 8954
    Prev
    Next
    INDEX
    Explanations

    initialize, allocate memory

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

    The neuron is highly responsive to words about initialization status—i.e. “initialize,” “initializing,” “uninitialized,” etc.

    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
    caya
    0.66
    专利
    0.56
     patented
    0.54
     Bolt
    0.53
    却是
    0.53
    削
    0.52
     ワンピース
    0.52
    অন্ত
    0.51
    ꯤ
    0.50
     Ginger
    0.50
    POSITIVE LOGITS
     repeated
    0.99
     repeat
    0.99
     repeating
    0.99
     repeats
    0.94
    repeat
    0.93
     повторя
    0.91
     repet
    0.91
     repetition
    0.91
    REPEAT
    0.91
     repetitive
    0.89
    Activations Density 0.008%

    No Known Activations