© 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. 90297
    Prev
    Next
    INDEX
    Explanations

    musical terms and prefixes

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

    The neuron is highly sensitive to tokens that appear in classical‐music article metadata and section headings—including performance/review language (e.g. “Music,” “Festival,” “finale,” “cadenza”), structural labels (e.g. “Structure,” “movements”), and associated numeric metadata (dates, movement counts, run‐time).

    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
     compléter
    -0.89
     Bigelow
    -0.85
    ımda
    -0.82
     complementos
    -0.80
    schirm
    -0.79
    composition
    -0.78
    plakat
    -0.76
     COMPOSITION
    -0.75
     arric
    -0.75
     stuk
    -0.74
    POSITIVE LOGITS
     Lys
    0.83
     markings
    0.76
     Boxes
    0.73
    issimo
    0.71
     Viva
    0.68
     con
    0.68
     camber
    0.68
     rit
    0.67
     фер
    0.67
     sempre
    0.66
    Activations Density 0.028%

    No Known Activations