© 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. 22-GEMMASCOPE-RES-131K
    7. 33944
    Prev
    Next
    INDEX
    Explanations

    course research

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

    This neuron detects structured section headings or labels (like “Office Hours:”, “Phone:”, “Required Course Content”) in academic or syllabus‐style documents.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    ֙
    -1.20
     oggi
    -1.09
    Эти
    -1.04
     Stadia
    -1.02
    还能
    -1.02
     bluse
    -1.02
    viamo
    -1.01
     творо
    -0.99
    жих
    -0.99
    licante
    -0.98
    POSITIVE LOGITS
     will
    1.89
     my
    1.61
     you
    1.61
     this
    1.50
     your
    1.45
     тази
    1.29
     youll
    1.12
     budete
    1.10
    看去
    1.09
     musíte
    1.09
    Activations Density 0.008%

    No Known Activations