SciReplicate-Bench 🏆 : a benchmark for NLP papers replication. lnkd.in/dpiVfn-8 Overthinking 🧠and lazy tool 🔨🪜recall is observed. Very proud to contribute to the work led by Yanzheng Xiang KCL NLP Stay tuned for our released data and code!

We are offering a fully-funded PhD studentship focused on using LLMs for Educational Assessment at KCL Informatics, in collaboration with AQA. Deadline to apply: May 31, 2025. kcl.ac.uk/study-legacy/f…

Congratulations to Siya Siya with 3 papers accepted by ACL 2025! 🎉 Check out these amazing works on LLM hallucination, Theory of Mind Reasoning, and AI4Edu: arxiv.org/abs/2503.01670

🎉 Excited to announce that the KclNLP group has 3 papers accepted at #ICML2025 and 15 papers accepted to #ACL2025! 👏 Huge congratulations to all the authors—stay tuned for more details! KCL Informatics KCL Natural, Mathematical & Engineering Sciences Human-Centred Computing @ King's College London Yulan He

📢 We’re offering a fully-funded PhD studentship in NLP at KCL! Join a vibrant research community and work on cutting-edge topics with leading experts. 🗓️ Deadline: May 31, 2025 🔗 Find more details at kcl.ac.uk/study-legacy/f… Yulan He

🔥 Chain-of-thought (CoT) reasoning in natural language is powerful—but inefficient. What if LLMs could reason in a compact continuous space instead? 🚀 Introducing CODI, our new paper on post-training LLMs for continuous reasoning via self-distillation. 💡TL;DR: CODI

SciReplicate-Bench has been accepted to Conference on Language Modeling 🥳 See you in Montreal! 🇨🇦 Website: xyzcs.github.io/scireplicate.g… (Data and code will be released later this month.)

🎉 Excited to share that our team has three papers accepted to ICML 2025, each exploring a different angle of how machines reason, from meta-reasoning to sparse and latent representations! 🧠 Spotlight (July 15) "Soft Reasoning: Navigating Solution Spaces in Large Language

ICYM Informatics' Yulan He is developing a pioneering AI virtual assistant designed to help examiners mark exam papers with greater accuracy and consistency. 🤖📝 She spoke with The Telegraph about how AI could help reduce the workload for teachers.⬇️ telegraph.co.uk/news/2025/07/1…

Look forward to starting this exciting journey in collaboration with Cesare Aloisi at AQA. KCL Informatics

Steer LLM reasoning using a single continuous token, without relying on SFT or RL. Check out our #ICML2025 poster E-2024 from the KCL NLP group: Soft Reasoning: Navigating Solution Spaces in Large Language Models through Controlled Embedding Exploration. Paper:

🚀 The KCL NLP Group is heading to #ACL2025 We’re excited to present our latest work on LLM reasoning, interpretability, personalization, and many other topics. 📍 Catch us in Vienna — stop by, say hi, and let’s talk!

🎉 It’s Day 1 at #ACL2025! The KCL NLP Group is kicking off the conference with a packed lineup of exciting posters. 📍Come find us — we’d love to connect! 🧠📄👇 Details in the image.

How do we tailor LLM according to user profiles and make them more personalized? Postdoc researcher Linhai Linhai Zhang is giving a poster presentation of his work on LLM personalization at Hall X5 51 right now. Come and check it out! #ACL2025NLP

📌 Day 2 at #ACL2025 — time for more posters! 🖼️ Catch us at today’s poster sessions — we’re excited to share ideas and connect! 📷 Full schedule in the image

🎤 Day 3 at #ACL2025! We’re wrapping up strong — more presentations from the KCL NLP Group. 🧠 Join us at today’s sessions to chat about our latest findings! 📷 Full schedule in the image.