© 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. 9379
    Prev
    Next
    INDEX
    Explanations

    beat, lose, pound, develop

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

    The neuron fires on vivid descriptions of loud, physical conflict scenes—especially onomatopoeic or high-intensity action words (bang, wrestling, strikes, cheers, yells, 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
    颞
    0.68
     creo
    0.59
    工程师
    0.59
    䀓
    0.59
    믈
    0.59
    odata
    0.58
    单个
    0.58
    夎
    0.58
     होतात
    0.57
    automate
    0.57
    POSITIVE LOGITS
     вследствие
    0.89
     Furthermore
    0.85
     Qing
    0.83
     Hongkong
    0.83
     Jiangsu
    0.83
    Furthermore
    0.82
    Alright
    0.82
     Zhao
    0.80
     Weibo
    0.80
     Xiao
    0.79
    Activations Density 0.044%

    No Known Activations