© 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. 24-GEMMASCOPE-2-RES-16K
    7. 15209
    Prev
    Next
    INDEX
    Explanations

    furthermore, also

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

    The neuron selectively fires on sentence-initial discourse-linking adverbial phrases (e.g. “Further,” “In addition,” “Moreover,”) that mark transitions or additive connections.

    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_24_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
     conduits
    0.82
     그러나
    0.75
    یت
    0.74
    ination
    0.73
    ிறது
    0.73
    klassen
    0.73
     marie
    0.73
     muck
    0.73
     tehát
    0.72
     transform
    0.71
    POSITIVE LOGITS
     هناك
    0.88
    要知道
    0.82
     اگر
    0.78
     ارد
    0.75
    સ
    0.75
     상당
    0.74
    しても
    0.74
    モニター
    0.74
     jeśli
    0.73
     جبکہ
    0.73
    Activations Density 0.102%

    No Known Activations