© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 43661
    Prev
    Next
    INDEX
    Explanations

    contrastive conjunctions

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

    This neuron detects concessive discourse markers—subordinating conjunctions like “while” and “although” that signal contrast.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     those
    -1.59
     it
    -1.59
     Vorteil
    -1.52
     However
    -1.50
    يمه
    -1.49
     rique
    -1.48
     kuitenkin
    -1.48
     ollut
    -1.46
     Because
    -1.43
     distancias
    -1.41
    POSITIVE LOGITS
     kila
    1.36
     系列
    1.24
     hispan
    1.22
     globale
    1.21
     rio
    1.21
     glau
    1.20
     appels
    1.20
     jd
    1.19
     tsi
    1.19
    woch
    1.19
    Activations Density 0.027%

    No Known Activations