© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 31158
    Prev
    Next
    INDEX
    Explanations

    ordinal grade levels

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

    The neuron detects numeric mentions of school grades (e.g., “fifth grade,” “4th grade,” etc.).

    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
     teens
    -1.02
     tenth
    -0.99
     teenage
    -0.93
     eleventh
    -0.88
     teenagers
    -0.87
     ninth
    -0.85
     bah
    -0.84
     fourteenth
    -0.79
    MHz
    -0.78
     fourteen
    -0.78
    POSITIVE LOGITS
     grade
    1.41
     grades
    1.11
     Grade
    1.06
    Grade
    1.01
    grade
    0.97
     graders
    0.87
    grades
    0.87
    GRADE
    0.85
    fourth
    0.73
     grader
    0.71
    Activations Density 0.023%

    No Known Activations