© 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. 107551
    Prev
    Next
    INDEX
    Explanations

    palmitoyl, Pal, Palmyra, Palmdale, names

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

    proper names, especially those with unique or ethnic spellings.

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

    The neuron strongly activates on fragments of proper names (and other tokens) beginning with “Pal.”

    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
    duke
    -0.79
    schaft
    -0.79
     Mistakes
    -0.77
    裘
    -0.77
    ultad
    -0.73
    Desen
    -0.72
    asjoner
    -0.71
    补贴
    -0.71
    বে
    -0.69
    сос
    -0.69
    POSITIVE LOGITS
     기
    0.87
    therly
    0.80
    cles
    0.78
     pal
    0.76
     favoured
    0.76
    fante
    0.74
     RATING
    0.74
    久了
    0.73
    Pal
    0.73
    Xan
    0.72
    Activations Density 0.026%

    No Known Activations