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

    Neuronpedia

    INDEX
    Explanations

    recommend or suggest

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

    references to the speaker or addressee in a personal, conversational tone (first- and second-person address)

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     malfunctions
    0.35
     ablation
    0.33
     isotropy
    0.32
     metabolism
    0.31
     appliances
    0.30
     malfunction
    0.30
     rotting
    0.30
     denaturation
    0.30
     servings
    0.30
     ਆਪਣ
    0.29
    POSITIVE LOGITS
     recommend
    0.50
     suggest
    0.47
     sugiere
    0.47
     recommande
    0.46
     recommends
    0.45
    建議
    0.45
     empfehlen
    0.45
     recommending
    0.45
    建议
    0.44
     suggested
    0.42
    Act Density 0.288%

    No Known Activations