© 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. 22-GEMMASCOPE-RES-131K
    7. 71269
    Prev
    Next
    INDEX
    Explanations

    unexpected contrasts

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

    The neuron primarily detects the adverb “even” when used as a contrastive or emphatic marker.

    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
     they
    -1.22
    艤
    -1.08
     lika
    -1.00
    ches
    -0.97
    pane
    -0.94
     gimnas
    -0.94
     apartament
    -0.94
    ísimas
    -0.93
     آن‌ها
    -0.92
    tea
    -0.91
    POSITIVE LOGITS
    展示
    1.05
    Можно
    1.04
    才可以
    0.98
     نہیں
    0.93
     cannot
    0.93
    loir
    0.92
    pecabe
    0.91
     galima
    0.90
     bubuk
    0.90
     puede
    0.89
    Activations Density 0.020%

    No Known Activations