© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 34556
    Prev
    Next
    INDEX
    Explanations

    or, and, as

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

    The neuron specifically detects the coordinating conjunction “or” when it’s used to link alternatives or list items.

    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
    ğini
    -1.30
     tiroirs
    -1.15
    łeś
    -1.15
    taient
    -1.14
     encontrou
    -1.13
    ꪻ
    -1.13
    模样
    -1.13
    ğine
    -1.13
     этого
    -1.12
    łby
    -1.12
    POSITIVE LOGITS
     to
    2.39
     as
    2.03
     purposes
    1.95
     for
    1.86
     для
    1.23
     Purposes
    1.23
     needs
    1.20
    purposes
    1.16
     just
    1.16
    เพื่อ
    1.09
    Activations Density 0.129%

    No Known Activations