© 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. 34-GEMMASCOPE-RES-131K
    7. 7004
    Prev
    Next
    INDEX
    Explanations

    conjunctions initiating clauses

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

    The neuron strongly activates on short function words that serve as discourse markers—especially contrastive or continuing conjunctions and modals (e.g. “but,” “so,” “can,” “never”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     bekym
    -0.98
    antaranya
    -0.94
     kampen
    -0.92
     PÉ
    -0.92
     tillegg
    -0.91
    ことになる
    -0.91
    жие
    -0.91
    thée
    -0.90
     uska
    -0.90
     urat
    -0.90
    POSITIVE LOGITS
     with
    3.64
    With
    1.94
     With
    1.92
     through
    1.87
     by
    1.84
     achievable
    1.74
     if
    1.73
    只要
    1.56
     WITH
    1.50
    Through
    1.48
    Activations Density 0.052%

    No Known Activations