© 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. 10-GEMMASCOPE-RES-131K
    4. 118897
    Prev
    Next
    INDEX
    Explanations

    corruption and disruption

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

    The neuron triggers on the subword “rupt,” i.e. in words like “corrupt,” “disrupt,” “endocrine‐disrupting,” etc.

    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
    厹
    -2.55
    伥
    -2.44
    actéristi
    -2.33
    ierenden
    -2.28
     echte
    -2.22
     quintessential
    -2.19
    噲
    -2.17
    豨
    -2.17
     colorida
    -2.11
    神色
    -2.11
    POSITIVE LOGITS
    '
    3.36
    \
    2.78
    其妙
    2.75
    while
    2.44
    with
    2.44
    🤠
    2.39
    *
    2.39
    剧照
    2.36
     seems
    2.22
    的很
    2.17
    Activations Density 0.006%

    No Known Activations