© 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. 82305
    Prev
    Next
    INDEX
    Explanations

    island and mainland names

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

    The neuron strongly activates on words and phrases from Shakespearean passages (e.g. character names, quoted lines, and related commentary) and not on non-Shakespeare text.

    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
    Datuak
    -0.88
     Galla
    -0.88
     знают
    -0.84
    liage
    -0.82
    āju
    -0.81
     bucket
    -0.79
     Sutter
    -0.78
    攤
    -0.78
     Crusader
    -0.77
    Itoa
    -0.77
    POSITIVE LOGITS
     Trin
    0.92
     Pros
    0.88
     island
    0.86
    Trin
    0.79
     острова
    0.78
     eiland
    0.74
     Miranda
    0.73
     mainland
    0.72
     shipwreck
    0.71
     Insel
    0.71
    Activations Density 0.008%

    No Known Activations