© 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-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 34889
    Prev
    Next
    INDEX
    Explanations

    code snippets and programming terms

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

    The neuron detects the characteristic tokens of JavaScript eval‐based obfuscation (e.g. the “.split('|')” call, the numeric “0” argument, and the surrounding packed eval/function structure).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Grill
    -0.85
     arkas
    -0.84
     METHODOLOGY
    -0.84
    이드
    -0.84
    Toaster
    -0.82
    某个
    -0.81
     допомогти
    -0.81
     attraction
    -0.81
     poch
    -0.79
    меров
    -0.79
    POSITIVE LOGITS
     Odd
    0.77
    propto
    0.76
    EMU
    0.71
    elor
    0.71
     yester
    0.69
    eterminate
    0.69
    もしろ
    0.69
     discutir
    0.68
     impra
    0.68
    visning
    0.66
    Activations Density 0.001%

    No Known Activations