Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    references to delicate handling and care in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Fragile or easily damaged things

    np_acts-logits-general · gemini-2.0-flash

    sensitive to damage or destruction

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     FetchType
    -0.58
    bluza
    -0.49
    Rhestr
    -0.48
     FishBase
    -0.46
    klärt
    -0.45
     rivalry
    -0.44
    bige
    -0.44
     adop
    -0.44
    agr
    -0.44
    iteracy
    -0.44
    POSITIVE LOGITS
     delicate
    0.69
     fragile
    0.66
     sensitive
    0.57
     frág
    0.51
     delicado
    0.49
    繊細
    0.45
     delikat
    0.45
     undamaged
    0.45
     damage
    0.45
     precious
    0.43
    Act Density 0.350%

    No Known Activations