© 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. 13299
    Prev
    Next
    INDEX
    Explanations

    bullying and aggression

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

    The neuron strongly activates on mentions of bullying (bullied, bully, victim, stop the bully, etc.) and related guidance or advice about it.

    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
     супру
    0.79
     पती
    0.76
     suami
    0.65
    婴
    0.64
     путеше
    0.63
     marido
    0.63
    🤰
    0.62
     vợ
    0.61
    榻
    0.61
    hapi
    0.61
    POSITIVE LOGITS
     bullies
    2.09
     bullying
    2.02
     bully
    1.98
     bullied
    1.92
     playground
    1.68
    bullying
    1.63
     Playground
    1.42
     classmate
    1.39
     classmates
    1.38
     school
    1.37
    Activations Density 0.232%

    No Known Activations