© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 101636
    Prev
    Next
    INDEX
    Explanations

    condition or build directive

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

    conditional compilation directives and build tags in code (like "if", "endif", "build").

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

    This neuron detects the “if … endif” markers used in HTML conditional comments (e.g. <!--[if IE]> … <![endif]-->).

    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
    ↵↵
    -1.77
     "
    -1.66
    起到
    -1.59
    Rape
    -1.53
    ↵
    -1.50
     jednou
    -1.50
     because
    -1.49
     ilumina
    -1.45
     acabado
    -1.44
     predomin
    -1.38
    POSITIVE LOGITS
     pubblicato
    1.66
     regista
    1.62
    甞
    1.59
     jorge
    1.55
     寺
    1.54
     colombia
    1.54
     valentino
    1.53
     balcon
    1.53
     of
    1.52
     pedro
    1.52
    Activations Density 0.006%

    No Known Activations