© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 98961
    Prev
    Next
    INDEX
    Explanations

    code and function definitions

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

    The neuron detects the opening curly‐brace that begins a control‐flow block (e.g. after “for … {” or “if … {”) in Go code.

    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
     a
    -2.66
    肫
    -1.89
    ”.
    -1.81
    ”
    -1.81
    -
    -1.80
    大さじ
    -1.77
    ’
    -1.76
     étrange
    -1.70
     “
    -1.70
    了一個
    -1.68
    POSITIVE LOGITS
     quedará
    1.99
     obtend
    1.85
     этого
    1.80
    mės
    1.76
    fficking
    1.76
     diverso
    1.73
     וגם
    1.73
     exitoso
    1.72
    because
    1.71
     or
    1.70
    Activations Density 0.003%

    No Known Activations