© 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. 22-GEMMASCOPE-RES-131K
    7. 71693
    Prev
    Next
    INDEX
    Explanations

    course descriptions and objectives

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

    This neuron fires on the boilerplate language of academic course listings—phrases that state what a course “is designed to,” “will,” or “provides” for students (i.e. course objectives/outcomes).

    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
     there
    -1.16
    ればいい
    -1.11
     acabó
    -1.08
    secutive
    -1.07
     seleccione
    -1.06
    tiamo
    -1.06
     funcionalidad
    -1.06
    konstruktion
    -1.04
     these
    -1.03
     possíveis
    -1.02
    POSITIVE LOGITS
     candidat
    1.27
    主要是
    1.26
     dianteiro
    1.23
     is
    1.21
    əl
    1.19
     rigide
    1.16
    AfterViewInit
    1.16
    พาะ
    1.14
     brasil
    1.13
     trase
    1.11
    Activations Density 0.018%

    No Known Activations