© 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. 84031
    Prev
    Next
    INDEX
    Explanations

    crystal structure

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

    The neuron activates strongly on passages describing molecular or macromolecular structures—especially titles or statements of “crystal structure” and related structural biology determinations.

    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
    }}]{
    -0.77
     Jojo
    -0.76
    Jojo
    -0.75
    idents
    -0.71
     dorm
    -0.70
    pover
    -0.70
     creations
    -0.70
    plots
    -0.69
    vestimento
    -0.68
    #:
    -0.68
    POSITIVE LOGITS
     structure
    2.11
     structures
    1.95
     Structures
    1.71
     Structure
    1.70
    Structure
    1.68
     crystal
    1.65
     STRUCTURES
    1.59
    structure
    1.55
    Structures
    1.53
    PDB
    1.48
    Activations Density 0.173%

    No Known Activations