© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 87414
    Prev
    Next
    INDEX
    Explanations

    labyrinth and maze

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

    The neuron activates on words referring to mazes or labyrinths (e.g. “labyrinth,” “maze,” and their variants).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    栳
    -2.61
    釭
    -2.30
     zorgt
    -2.27
    雠
    -2.27
    葙
    -2.23
    entown
    -2.20
    ⡦
    -2.20
    慊
    -2.19
    榈
    -2.14
    祧
    -2.14
    POSITIVE LOGITS
     by
    2.14
     but
    2.13
     "
    2.06
    ساء
    1.87
     not
    1.83
     for
    1.83
    .'
    1.78
    不对
    1.78
    )$
    1.77
    ৄ
    1.77
    Activations Density 0.007%

    No Known Activations