© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 8016
    Prev
    Next
    INDEX
    Explanations

    comparing, comparison, comparisons

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

    The neuron is on the lookout for explicit comparison or contrast cues (e.g. “compare,” “comparing,” “contrast,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ುವ
    0.54
     Buildable
    0.53
     تنفيذ
    0.53
    úan
    0.51
    充滿
    0.49
     beinhaltet
    0.49
     സംഭ
    0.49
    绪
    0.49
    謨
    0.48
    ozer
    0.48
    POSITIVE LOGITS
     comparison
    4.28
     comparisons
    4.16
    comparison
    4.06
    Comparison
    3.97
     comparing
    3.95
     Comparison
    3.95
     Comparisons
    3.91
     срав
    3.83
    Comparing
    3.77
     comparaison
    3.77
    Activations Density 0.549%

    No Known Activations