© 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. 2327
    Prev
    Next
    INDEX
    Explanations

    code and file paths

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

    The neuron activates on actual code tokens (keywords, literals, identifiers) rather than on comments or prose.

    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
     when
    -0.75
     çocu
    -0.74
     Antônio
    -0.74
    ológicas
    -0.73
     ALWAYS
    -0.73
     cumbersome
    -0.72
     Although
    -0.71
     membres
    -0.71
     Taking
    -0.70
     although
    -0.70
    POSITIVE LOGITS
    ͻ
    0.80
     иссле
    0.80
     associations
    0.78
     защи
    0.77
     Selenium
    0.75
    yourself
    0.74
     transcrip
    0.74
     Hei
    0.74
     PZ
    0.74
    Ȓ
    0.74
    Activations Density 0.001%

    No Known Activations