© 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. 80673
    Prev
    Next
    INDEX
    Explanations

    locations and descriptions

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

    The neuron fires on concrete, factual tokens – things like named entities (place names, organization names), numeric values (dates, measurements, coordinates), and technical constants – essentially the bits that carry “hard” data rather than function words.

    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
     酷
    -0.67
    Unit
    -0.67
     pintu
    -0.65
    Δια
    -0.65
     unit
    -0.65
     mehrerer
    -0.63
    Tiny
    -0.63
    livé
    -0.63
     celana
    -0.63
     spécialement
    -0.62
    POSITIVE LOGITS
    alais
    0.81
    стка
    0.78
    ForResult
    0.76
     uniformes
    0.72
    circling
    0.71
    zys
    0.70
    moveToFirst
    0.70
    chuga
    0.69
     Cô
    0.68
     cartilage
    0.67
    Activations Density 0.073%

    No Known Activations