© 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. 116695
    Prev
    Next
    INDEX
    Explanations

    marker retainer

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

    The neuron detects product‐style tokens—brand names, model numbers, or alphanumeric product identifiers (e.g. “Promarkers,” “Copic,” “92”).

    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
     pencil
    -1.02
     Pencil
    -0.99
    Pencil
    -0.97
    pencil
    -0.92
     каранда
    -0.83
    hyp
    -0.79
    鉛筆
    -0.77
    HasValue
    -0.76
     pencils
    -0.75
     кӀош
    -0.75
    POSITIVE LOGITS
     marker
    2.34
     markers
    2.23
     Marker
    1.87
     Markers
    1.82
    marker
    1.82
    Marker
    1.76
    Markers
    1.76
    markers
    1.62
    MARKER
    1.55
     marcadores
    1.48
    Activations Density 0.027%

    No Known Activations