© 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. 12503
    Prev
    Next
    INDEX
    Explanations

    Saving Survival

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

    The neuron fires strongly on instruction‐style action words—especially single verbs that kick off or headline steps (e.g. “changing,” “decompiling,” “rip,” “lean away,” “filter,” “linking,” “vibe”)—marking the start of procedural or how-to directions.

    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
    ir
    0.82
     ST
    0.76
    h
    0.73
    un
    0.72
     C
    0.71
     Tunis
    0.70
     Nella
    0.69
     IST
    0.68
     Nelson
    0.68
    al
    0.68
    POSITIVE LOGITS
     oxidizing
    0.89
     lojas
    0.87
     wynik
    0.86
    দিগকে
    0.85
     ګټ
    0.84
    ты
    0.83
    єд
    0.83
    চ্ছিলাম
    0.83
     사항
    0.82
     połą
    0.82
    Activations Density 0.000%

    No Known Activations