From Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/1704.07926/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTFrom Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood — mentioned by ke1vinhttps://news.ycombinator.com/item?id=14381315urn:hackerpapers:mention:1704.07926:14381315Sat, 20 May 2017 06:43:29 GMT<p>When you only provide the system with a few examples, there are many possible transformations which satisfy the examples. One way to eliminate this ambiguity is to also provide a natural language description of what you want, e.g. &quot;remove the spaces&quot;. In the natural language processing community, we call this semantic parsing. But sometimes the semantic parser can misinterpret the language too and generate a program which still &quot;cheats&quot; in the same manner as you described. We call these &quot;spurious programs&quot;. Shameless plug-- my group has been working on how to deal with these spurious…</p><p><a href="https://hackerpaper.kaipfister.de/p/1704.07926/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.AI