© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 2819
    Prev
    Next
    INDEX
    Explanations

    subtitles and transcripts

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

    The neuron fires on mentions of subtitle/captioning or dubbed‐language tracks (e.g. “subtitles,” “captioned,” “dubbed English/French/Spanish,” 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
    Registre
    0.59
    Knight
    0.57
    ImageBox
    0.57
    )^
    0.56
    ;&#
    0.56
    серва
    0.56
    FACT
    0.55
     Hobby
    0.55
    USP
    0.55
    LinkedList
    0.54
    POSITIVE LOGITS
     captions
    0.98
    字幕
    0.96
     caption
    0.86
     subtitles
    0.83
     Caption
    0.72
     transcripts
    0.68
    텅
    0.67
    caption
    0.66
     captioned
    0.64
     timing
    0.64
    Activations Density 0.021%

    No Known Activations