© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 113729
    Prev
    Next
    INDEX
    Explanations

    SIMD instructions and vectorization

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

    SIMD (Single Instruction Multiple Data) and vectorization-related code, particularly in assembly, C/C++, and low-level programming contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires on tokens naming CPU architectures or SIMD/vector‐extension features (e.g. NEON, SVE, SSE2, AVX2, MMX, SIMD, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     could
    -1.83
     had
    -1.68
    un
    -1.48
     providing
    -1.48
     provides
    -1.45
     provide
    -1.45
     indicating
    -1.45
     Depending
    -1.43
     gave
    -1.42
     began
    -1.42
    POSITIVE LOGITS
    媗
    1.83
     tunik
    1.70
     frequentemente
    1.59
    𖧵
    1.53
    Compte
    1.51
    龊
    1.43
     etik
    1.41
     způ
    1.41
    Jiang
    1.39
    nahm
    1.39
    Activations Density 0.017%

    No Known Activations