Post
312
Inverse IFEval 🔥New benchmark from Bytedance & MAP
m-a-p/Inverse_IFEval
Inverse IFEval: Can LLMs Unlearn Stubborn Training Conventions to Follow Real Instructions? (2509.04292)
Testing LLMs on their ability to override biases & follow adversarial instructions.
✨ 8 challenge types
✨ 1,012 CN/EN Qs across 23 domains
✨ Human-in-the-loop + LLM-as-a-Judge
m-a-p/Inverse_IFEval
Inverse IFEval: Can LLMs Unlearn Stubborn Training Conventions to Follow Real Instructions? (2509.04292)
Testing LLMs on their ability to override biases & follow adversarial instructions.
✨ 8 challenge types
✨ 1,012 CN/EN Qs across 23 domains
✨ Human-in-the-loop + LLM-as-a-Judge