© 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. 82911
    Prev
    Next
    INDEX
    Explanations

    annual fund pledge

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

    The neuron selectively fires on metadata‐style tokens—particularly section headings, field names, classification codes, and other label‐like words (e.g. MSC codes, “Brand,” “Abstract,” “Title,” “Event,” 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_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
    Jas
    -0.84
     placée
    -0.83
     czuj
    -0.75
    віта
    -0.71
     iulie
    -0.69
     zobaczyć
    -0.68
    怎樣
    -0.66
    dedores
    -0.66
    below
    -0.66
     Bounty
    -0.66
    POSITIVE LOGITS
     khách
    0.66
    敷
    0.65
    paper
    0.65
    PRESSED
    0.64
    rijk
    0.64
     PIL
    0.64
    towanie
    0.63
    ättern
    0.63
     sella
    0.63
    ับ
    0.62
    Activations Density 0.074%

    No Known Activations