Thom Wolf 谈 LLM 短板：心智理论与"知识的诅咒"
What's still very hard: theory of mind. The world, the voices and the mechanics were great. Writing ...
Transformers 作者本人说模型讲故事会"替玩家全知"，NPC 反应错乱，还聊到为什么教学难用 RL 训练，观点挺扎心。
NLNet/早期 Transformers 作者 Thom Wolf 用 AI 生成互动故事的经历指出，模型在画面、声音和机制上表现不错，但写作时假设玩家知道得比实际多，NPC 会回应从未发生的事情。他认为 AI 导师类产品需要建模学习者的心智状态，而这正是当前 LLM 的弱项。他补充说，教学任务缺乏可验证奖励，在现有 RL 训练范式下难以规模化。
What's still very hard: theory of mind. The world, the voices and the mechanics were great. Writing ...
What's still very hard: theory of mind. The world, the voices and the mechanics were great. Writing the story was the weak part: the model assumes the player knows more than they do, and its NPC sometimes react in strange way or to things that never happened. A Diamond Age's Primer (or any real AI tutor fwiw) will have to model the learner, and that is where it still falls short in a default setting imo. I've actually started exploring benchmarks on this and it's a hard task for LLMs right now. They are prone to the "Curse of knowledge". Teaching humans is also not very much a Verifiable Reward, so it's hard to scale within the current training paradigm. 💬 1 🔄 0 ❤️ 0 👀 705 📊 1 ⚡
