© 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. 65515
    Prev
    Next
    INDEX
    Explanations

    pirate, viking, immortal, peptide

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

    references to pirates, Vikings, immortality, or related maritime or legendary concepts.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron fires on mentions of “pirate” (and closely related seafaring-raider labels like “space pirate,” with lesser responses to other similar proper-noun labels such as “Viking”).

    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
    \
    -3.00
    錞
    -2.34
    "
    -2.17
     eatery
    -2.13
     assis
    -2.03
     современных
    -2.03
     jarring
    -2.03
     моей
    -2.02
     personnalité
    -2.00
     aching
    -1.95
    POSITIVE LOGITS
     ‘‘
    2.52
    
    2.47
    地说道
    2.47
    </h1>
    2.45
    蹌
    2.39
    )』
    2.33
     {
    
    2.33
    isation
    2.27
     jaja
    2.22
    𓋼
    2.20
    Activations Density 0.008%

    No Known Activations