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

    Neuronpedia

    INDEX
    Explanations

    sensitive themes and difficult emotions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     hints
    0.38
     sneaky
    0.36
     snappy
    0.36
     phishing
    0.35
     deforestation
    0.35
     tweaking
    0.35
     alphabetical
    0.34
     overarching
    0.34
     flashbacks
    0.34
     hikes
    0.34
    POSITIVE LOGITS
     dynamics
    0.41
    Dynamics
    0.40
     realities
    0.38
     אות
    0.36
    feira
    0.35
    liš
    0.35
    dynamics
    0.35
    结论
    0.33
    ਰੇ
    0.33
     proportions
    0.33
    Act Density 0.029%

    No Known Activations