© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 32456
    Prev
    Next
    INDEX
    Explanations

    performs rituals and ceremonies

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

    The neuron strongly responds to fire‐related roots (e.g. “burning,” “pyre,” “py–” fragments), so it’s detecting terms about burning or fire.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    Pseudo
    -0.88
    يتر
    -0.78
    石川
    -0.74
     Squash
    -0.73
    льних
    -0.72
     Pseudo
    -0.71
    Реги
    -0.71
     Tern
    -0.70
     extrinsic
    -0.70
     gydy
    -0.70
    POSITIVE LOGITS
     ceremony
    1.25
     performs
    1.10
     ritual
    1.09
     rituals
    1.08
     ceremonies
    1.04
    Perform
    1.03
     perform
    1.01
    perform
    0.98
     performed
    0.98
     performer
    0.96
    Activations Density 0.011%

    No Known Activations