© 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. 11935
    Prev
    Next
    INDEX
    Explanations

    door and gate related words

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

    The neuron detects mentions of doors or gates and related action verbs (e.g., “open,” “swing,” “slam”).

    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
     đốt
    0.61
    띄
    0.61
    화
    0.56
     refroidissement
    0.53
     xuống
    0.53
    シャー
    0.49
     sillonné
    0.49
     sulfonic
    0.49
     BOARDSIZE
    0.49
    گی
    0.48
    POSITIVE LOGITS
     door
    2.64
     doors
    2.52
    门
    2.52
    Door
    2.50
     Door
    2.44
    門
    2.41
    door
    2.39
     дверь
    2.34
     pintu
    2.31
     puerta
    2.30
    Activations Density 0.396%

    No Known Activations