© 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. 34-GEMMASCOPE-RES-131K
    7. 91631
    Prev
    Next
    INDEX
    Explanations

    German and descriptive words

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

    The neuron detects adversative discourse markers and contrastive cue words (e.g. “however,” “but,” “also,” etc.).

    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
    腓
    -0.95
     my
    -0.93
     then
    -0.92
    沿い
    -0.91
    🥎
    -0.88
     angry
    -0.87
    依然
    -0.86
     exploring
    -0.85
     نکن
    -0.85
    ds
    -0.84
    POSITIVE LOGITS
     Accessoires
    1.10
    ㅡ
    1.08
     deem
    1.07
     otoño
    1.05
    kamera
    1.03
     anhänger
    1.03
     majority
    1.02
    Hersteller
    1.00
    Viel
    1.00
     albeit
    0.99
    Activations Density 0.007%

    No Known Activations