“Only tangentially related, but I think the mid/late game is a much more interesting/useful/hard avenue for chess tools. Chess.com's opening explorer is pretty good, which it seems like this is kind of trying to compete against. What I struggle with is the post game review, where I make a dubious move but I don't actually understand the reason why it's b…”
Improving Chess Commentaries by Combining Language Models with Symbolic Reasoning Engines
Despite many recent advancements in language modeling, state-of-the-art language models lack grounding in the real world and struggle with tasks involving complex reasoning. Meanwhile, advances in the symbolic reasoning capabilities of AI have led to systems that outperform humans in games like chess and Go (Silver et al., 2018). Chess commentary provides an interesting domain for bridging these two fields of research, as it requires reasoning over a complex board state and providing analyses in natural language. In this work we demonstrate how to combine symbolic reasoning engines with controllable language models to generate chess commentaries. We conduct experiments to demonstrate that our approach generates commentaries that are preferred by human judges over previous baselines.
候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。
HN 书评
推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及- Improving Chess Commentaries by Combining [LLMs] with Symbolic Reasoning Engines ↗eternalban · 2022-12-19 · 1 分