© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 1029
    Prev
    Next
    INDEX
    Explanations

    conjunctions of contrast: although, though

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

    This neuron is specialized to detect numeric expressions—dates, percentages, counts, and other multi-digit figures—in the text.

    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
    、
    0.66
     の
    0.57
    ”、“
    0.56
    を楽し
    0.55
    بھی
    0.55
     を
    0.54
     یې
    0.53
    の
    0.53
     この
    0.52
    Untuk
    0.52
    POSITIVE LOGITS
     although
    2.03
    although
    1.79
     though
    1.78
     aunque
    1.74
     according
    1.71
    though
    1.63
     obwohl
    1.59
    according
    1.58
     según
    1.38
    aunque
    1.33
    Activations Density 0.428%

    No Known Activations