© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 2627
    Prev
    Next
    INDEX
    Explanations

    prevent, escape, survive

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

    The neuron is sensitive to key content words in instructional or warning contexts—terms used when giving advice, describing dangers, or outlining how to proceed.

    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
     dissatisfaction
    0.62
    かね
    0.56
     unhappiness
    0.55
    鞭
    0.54
    嫉
    0.54
    追求
    0.54
    opharm
    0.54
     Simultaneous
    0.53
    ➟
    0.53
     Excellence
    0.52
    POSITIVE LOGITS
     survived
    3.22
     survive
    3.13
     survives
    2.95
     survival
    2.94
     surviving
    2.72
     sobrevivir
    2.70
     sobreviv
    2.67
    survival
    2.64
     Survive
    2.64
     supervivencia
    2.58
    Activations Density 0.793%

    No Known Activations