© 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-3-27B-IT
    3. 31-GEMMASCOPE-2-TRANSCODER-262K
    4. 33999
    Prev
    Next
    INDEX
    Explanations

    This neuron is associated with code structures for input/output and conditional control flow.Let's refine this based on the prompt's request for a phrase *without* mentioning "tokens" or "patterns," and aiming for conciseness and specificity.**TOKENS_AFTER_MAX_ACTIVATING_TOKEN**:`printf`, `if`, `std`, `printf`, `cout`, `//`, `scanf`, `//`, `printf`, `std`These are overwhelmingly C/C++ standard library keywords and functions, particularly for input (`scanf`, `cin`) and output (`printf`, `cout`), along with control flow (`if`) and namespaces (`std`).**TOP_POSITIVE_LOGITS**:`र्यंत`, `pgamma`, `deer`, `stranded`, `flamingo`, `विंडो`, `evole`, `FLY`, `ကြည့်`, `蟆`This list seems highly diverse and not directly related to programming constructs.**C/C++ I/O and control flow**

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_31_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     interactive
    0.39
     전시
    0.38
     kun
    0.37
     took
    0.37
    ams
    0.36
     exhibition
    0.34
     overall
    0.34
     latter
    0.34
     displays
    0.34
     tabular
    0.34
    POSITIVE LOGITS
    र्यंत
    0.42
    pgamma
    0.39
    deer
    0.38
    stranded
    0.37
    flamingo
    0.36
     विंडो
    0.36
    evole
    0.35
    FLY
    0.35
    ကြည့်
    0.35
    蟆
    0.35
    Activations Density 0.004%

    No Known Activations