© 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. 189
    Prev
    Next
    INDEX
    Explanations

    dialogue attribution verbs

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

    The neuron strongly fires on melodramatic or sensational narrative cues—especially exclamatory, emotionally charged phrases describing conflict or urgency.

    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
    有一些
    0.38
    াকেও
    0.38
    どのような
    0.37
    有哪些
    0.36
    也可以
    0.36
    可以使用
    0.35
    টিকে
    0.34
    用于
    0.34
     बनानी
    0.33
    不需要
    0.33
    POSITIVE LOGITS
     said
    1.06
     exclaimed
    1.06
     says
    1.00
    said
    0.88
    says
    0.87
     muttered
    0.86
     сказал
    0.85
     Says
    0.84
     replied
    0.83
     remarked
    0.80
    Activations Density 0.239%

    No Known Activations