© 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. 34-GEMMASCOPE-RES-131K
    7. 54286
    Prev
    Next
    INDEX
    Explanations

    fire prevention and fighters

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

    The neuron strongly activates on any occurrence of the substring “fire” (including “fire,” “fires,” “firefighter,” “fire-related,” etc.).

    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
    fire
    -1.14
     fires
    -1.10
     FIRE
    -1.05
     fire
    -1.05
     Fires
    -0.99
    fires
    -0.97
     flame
    -0.94
     пожар
    -0.93
     wildfires
    -0.93
    ticion
    -0.91
    POSITIVE LOGITS
     prevention
    1.04
     department
    0.94
     Prevention
    0.84
     從
    0.84
     science
    0.82
    SCENE
    0.82
     fighters
    0.80
     officials
    0.80
    bertson
    0.78
    eated
    0.77
    Activations Density 0.014%

    No Known Activations