© 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. 8024
    Prev
    Next
    INDEX
    Explanations

    a horror

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

    The main thing this neuron does is detect language related to horror or spooky themes.

    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
    🥙
    0.63
    授予
    0.61
     Doha
    0.59
     Approval
    0.59
     Cyrus
    0.59
    錦
    0.59
     Mahan
    0.58
    🚄
    0.58
     dynast
    0.57
    Bharat
    0.57
    POSITIVE LOGITS
     horror
    2.16
     spooky
    2.16
     Horror
    2.14
     creepy
    1.97
     eerie
    1.97
     Halloween
    1.94
    Halloween
    1.92
    horror
    1.87
     halloween
    1.86
     haunted
    1.85
    Activations Density 0.722%

    No Known Activations