© 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. 129875
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly responds to high-level function words and discourse markers—especially common prepositions, conjunctions, and clause-introductory particles (e.g. “In,” “As,” “To,” “by,” “of,” “All,” “like,” etc.) that frame or connect explanatory or comparative statements.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    phrases starting with prepositions

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     caoutchouc
    -1.27
    餡
    -1.22
     this
    -1.20
     Regione
    -1.16
    では
    -1.16
     riviera
    -1.16
     turist
    -1.14
     probleme
    -1.14
     Archiproducts
    -1.14
     litoral
    -1.13
    POSITIVE LOGITS
     I
    2.14
     my
    1.78
    為に
    1.35
    也只能
    1.21
    烔
    1.20
     Tamaño
    1.17
    他才
    1.13
    へ向
    1.11
     soweit
    1.10
    事を
    1.09
    Activations Density 0.108%

    No Known Activations