© 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. 34-GEMMASCOPE-RES-131K
    7. 80955
    Prev
    Next
    INDEX
    Explanations

    masked or masking

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

    The neuron is keyed to spotting all‐caps identifier tokens (uppercase letters, digits, or underscores) such as XML tags, version constants, and macro names.

    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
    Surrey
    -0.79
    ʊ
    -0.69
    htt
    -0.68
     DW
    -0.66
     Dreamer
    -0.66
    PIECE
    -0.66
     Surrey
    -0.65
     sunken
    -0.65
    丰
    -0.65
     должен
    -0.62
    POSITIVE LOGITS
    masked
    0.97
    masking
    0.90
     USE
    0.89
     masked
    0.88
    Masked
    0.87
     masking
    0.85
    astra
    0.83
     emerge
    0.83
    MASK
    0.79
     Masking
    0.79
    Activations Density 0.053%

    No Known Activations