© 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. 65747
    Prev
    Next
    INDEX
    Explanations

    metal models

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

    This neuron fires strongly on technical, domain-specific terms and jargon typical of scientific research writing.

    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
    镰
    -0.77
     успо
    -0.75
    Farma
    -0.75
    ள்ள
    -0.73
     velcro
    -0.72
    ITIS
    -0.71
    addValue
    -0.71
     Nights
    -0.71
     hula
    -0.70
     jacquard
    -0.69
    POSITIVE LOGITS
     models
    1.13
     metal
    0.98
     model
    0.89
     métal
    0.89
     delayed
    0.88
    Metal
    0.87
     yield
    0.85
    metal
    0.84
     Metal
    0.82
     metals
    0.82
    Activations Density 0.040%

    No Known Activations