https://jessyli.com Associate Professor, UT Austin Linguistics. Part of UT Computational Linguistics https://sites.utexas.edu/compling/ and UT NLP https://www.nlp.utexas.edu/https://bsky.app/profile/jessyjli.bsky.socialhttps://bsky.app/profile/jessyjli.bsky.social/post/3miz72rd5sc2pIf you ask the same question with different framing/phrasing, do language models change their answers? This is super important in medicine because different info can have real consequences! Check out this new work from @hyesunyun.bsky.social [contains quote post or other embedded content]08 Apr 2026 21:05 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3miz72rd5sc2phttps://bsky.app/profile/jessyjli.bsky.social/post/3mht7yj6h2c24Heading to #EACL2026! 🇲🇦 Friday 11a Poster Session 6: LMs struggle to perform inferences around discourse connectives https://aclanthology.org/2026.eacl-long.289/ Sunday 5p TeachingNLP workshop: new course on discourse+generation https://aclanthology.org/2026.teachingnlp-1.9/ w/ @kanishka.bsky.social + Daniel Brubaker24 Mar 2026 18:40 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3mht7yj6h2c24https://bsky.app/profile/jessyjli.bsky.social/post/3mgxnxsoxf225Want to know how well the models can brainstorm connections across different concepts? Super excited about @manyawadhwa.bsky.social’s work on measuring associative creativity! [contains quote post or other embedded content]13 Mar 2026 19:36 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3mgxnxsoxf225https://bsky.app/profile/jessyjli.bsky.social/post/3mgduhwj3422uCheck out our special theme: new missions for NLP research! [contains quote post or other embedded content]05 Mar 2026 22:39 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3mgduhwj3422uhttps://bsky.app/profile/jessyjli.bsky.social/post/3mdq5vag6h22vCheck out @asher-zheng.bsky.social's work on quantifying strategic language in dialogue, just appeared in the Dialogue and Discourse journal. We study non-cooperative moves that are subtle to capture, where modern AI still have trouble comprehending. Work w/ David_Beaver [contains quote post or other embedded content]31 Jan 2026 15:44 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3mdq5vag6h22vhttps://bsky.app/profile/jessyjli.bsky.social/post/3mcxfr2ix4c25🚨Be careful with LLMs when you ask health related questions -- even when the model relies on "evidence"! Kaijie's paper reveals a key weakness and the tricky balance between safety and faithfulness 👉 [contains quote post or other embedded content]21 Jan 2026 19:28 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3mcxfr2ix4c25https://bsky.app/profile/jessyjli.bsky.social/post/3m53kuvh6ok2rIncredibly honored to serve as #EMNLP 2026 Program Chair along with @sunipadev.bsky.social and Hung-yi Lee, and General Chair @andre-t-martins.bsky.social. Looking forward to Budapest!! (With thanks to Lisa Chuyuan Li who took this photo in Suzhou!) [contains quote post or other embedded content]08 Nov 2025 02:39 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3m53kuvh6ok2rhttps://bsky.app/profile/jessyjli.bsky.social/post/3m3dafaqprs2kThink your LLMs “understand” words like although/but/therefore? Think again! They perform at chance for making inferences from certain discourse connectives expressing concession [contains quote post or other embedded content]16 Oct 2025 17:02 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3m3dafaqprs2khttps://bsky.app/profile/jessyjli.bsky.social/post/3m34otwasdc2x🚨 Does your LLM really understand code -- or is it just really good at remembering it? We built **PLSemanticsBench** to find out. The results: a wild mix. ✅The Brilliant: Top reasoning models can execute complex, fuzzer-generated programs -- even with 5+ levels of nested loops! 🤯 ❌The Brittle: 🧵14 Oct 2025 02:32 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3m34otwasdc2xhttps://bsky.app/profile/jessyjli.bsky.social/post/3m2nhtp57fs2cWe’re hiring faculty as well! Happy to talk about it at COLM! [contains quote post or other embedded content]08 Oct 2025 01:17 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3m2nhtp57fs2chttps://bsky.app/profile/jessyjli.bsky.social/post/3m2jxovjhy22vOn my way to #COLM2025 🍁 Check out jessyli.com/colm2025 QUDsim: Discourse templates in LLM stories arxiv.org/abs/2504.09373 EvalAgent: retrieval-based eval targeting implicit criteria arxiv.org/abs/2504.15219 RoboInstruct: code generation for robotics with simulators arxiv.org/abs/2405.2017906 Oct 2025 15:50 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3m2jxovjhy22vhttps://bsky.app/profile/jessyjli.bsky.social/post/3m23b65uems2uAll of us (@kanishka.bsky.social @kmahowald.bsky.social and me) are looking for PhD students this cycle! If computational linguistics/NLP is your passion, join us at UT Austin! For my areas see jessyli.com [contains quote post or other embedded content]30 Sep 2025 19:30 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3m23b65uems2uhttps://bsky.app/profile/jessyjli.bsky.social/post/3lzotgp6wa224Can AI aid scientists amidst their own workflows, when they do not know step-by-step workflows and may not know, in advance, the kinds of scientific utility a visualization would bring? Check out @sebajoe.bsky.social’s feature on ✨AstroVisBench: [contains quote post or other embedded content]25 Sep 2025 20:52 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3lzotgp6wa224https://bsky.app/profile/jessyjli.bsky.social/post/3lwa7gnmwlc2fThe Echoes in AI paper showed quite the opposite with also a story continuation setup. Additionally, we present evidence that both *syntactic* and *discourse* diversity measures show strong homogenization that lexical and cosine used in this paper do not capture. [contains quote post or other embedded content]12 Aug 2025 21:01 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3lwa7gnmwlc2fhttps://bsky.app/profile/jessyjli.bsky.social/post/3lv2l6x7fr22hTuesday at #ACL2025: Jan will be presenting this from 4-5:30pm in x4/x5! Turns out content selection in LLMs are highly consistent with each other, but not so much with their own notion of importance or with human’s… [contains quote post or other embedded content]28 Jul 2025 21:51 +0000at://did:plc:mncmdjjmwqn6eptiaa2wuzdn/app.bsky.feed.post/3lv2l6x7fr22h