Andreas Vlachos
@vlachos_nlp
Followers
5K
Following
6K
Media
48
Statuses
1K
Professor in NLP/ML at @Cambridge_CL, member of the PaNLP group: https://t.co/CcOYgtiRTv, Fellow of @FitzwilliamColl, @ELLISforEurope member
Cambridge, UK
Joined June 2016
Honoured to be among those funded by this scheme! The pressure is on to deliver! If you know anyone looking for a postdoc or a PhD on the topic of fact checking in NLP/ML, point them my way!
Congratulations to Dr Andreas Vlachos on winning an ERC Consolidator Grant! 🎉🎉 #ERCCoG @ERC_Research #EUfunded
28
20
234
#EMNLP2025 #Cambridge 📣New Paper📣 Not all disagreement is noise! 🧐 We use semantic entropy to separate genuine human label variation from annotation errors, under the assumption that explanations for different plausible labels are not contradictory. https://t.co/ZVM1VJEoEe
0
2
7
#EMNLP2025 #Cambridge 📣New Paper📣 LLMs still stumble at planning with interdependent time constraints from real dialogue. We introduce TCP, a benchmark for deriving optimal schedules from naturalistic conversations under temporal constraints. ⛓️ https://t.co/R03rRn3PjN
1
1
9
It’s #EMNLP2025 week! The Cambridge NLIP group is presenting 17 papers across a wide range of topics.🚀 Working on similar problems as the Cambridge NLIP group? Come find our researchers throughout the conference!
1
2
17
Excited to share that our paper on automatically uncovering how NLP papers frame their research will be presented this week at the EMNLP 2025 main conference in Suzhou by @vlachos_nlp! 🎉 Paper, code and data below👇 @michael_sejr @nedjmaou
What are NLP papers really saying about the purpose and use of their models/datasets? 🤔 Who are they for? What problems do they solve? How are they used? We built a framework + tool to: (1) analyze framing trends across papers (2) help authors reflect on their own framing 🧵
1
3
12
If you are at Cambridge, join our NLIP seminar series Fri 24 Oct, 12:00 in SS03 (CS lab). Dr. Eyal Kolman (Microsoft) presents "Evaluation with LLMs - Theoretical and Practical insights". The talk covers LLM-as-judge, pitfalls, benchmarks, and tuning. 👉 https://t.co/52mcUdTcSy
1
3
9
Cambridge is recruiting Assistant/Associate Professors in Theoretical Computer Science. #TCS #AcademicJobs Application deadline: 15 December 2025. Contact me for informal inquiries. See more details and apply here: https://t.co/M7tHg24XZl
1
8
43
Co-organised with the great team: @akhtarmubashara @RamiAly Rui Cao (announcing the shared task) @Yulongchen1010 Oana Cocarascu @ZifengDing6 @ZhenyunDeng @ZhijiangG @IamArpitMittal @michael_sejr @j6mes Chenxi Whitehouse
0
1
7
remember to submit papers on topics related to fact-checking, including the previous shared tasks on AVERITEC, FEVEROUS and FEVER. Deadline 6th of October (ARR) and 19th of December (direct submissions).
1
0
5
Excited to announce the 9th @FEVERworkshop collocated with #EACL2026 in Morocco: https://t.co/4bGGntzZEZ with a new shared task on #factchecking image-text claims with evidence from the web using the AVerImaTeC dataset: https://t.co/5ugCkmUnji Deadline 2nd of December! And...
1
7
15
Excited to share that our work “Ev2R: Evaluating Evidence Retrieval in Automated Fact-Checking” has been accepted to #TACL. 🎉 This is joint work with my great collaborators @michael_sejr and @vlachos_nlp. Ev2R is a new evidence evaluation metric that combines reference-based
1
3
13
📢I will be giving a talk at the @FEVERworkshop at @aclmeeting on *Thursday at 2pm* Vienna time. Come by to hear about "Thoughts You Can Trust? Evaluating the Faithfulness of Model-Generated Explanations and Their Effects on Human Performance" https://t.co/Q0LPmthopB
4
11
53
The Cambridge NLIP Group is in Vienna for ACL 2025 @aclmeeting, presenting our work in the main conference and several workshops throughout the week. Come chat with us! #ACL2025 #ACL2025NLP #NLProc
0
4
17
@aclmeeting @vlachos_nlp And don't miss out on the FEVER Workshop (@FEVERworkshop) on Thursday (31.07)! This year, we again have exciting papers and speakers, and we will discuss the latest AVeriTeC shared task on efficient and reproducible systems.
0
1
4
At @aclmeeting? Today (Monday, 28.07) at the 11:00 poster session, I'll present my final PhD paper: "TabVer: Tabular Fact Verification with Natural Logic". Joint work with the amazing @vlachos_nlp. Drop by to say hi :) Paper link📚:
arxiv.org
Fact verification on tabular evidence incentivises the use of symbolic reasoning models where a logical form is constructed (e.g. a LISP-style program), providing greater verifiability than fully...
1
3
23
Looking forward to this year's edition! With great speakers: Ryan McDonald @yulanhe Vlad Niculae @anas_ant @raquel_dmg @annargrs @preslav_nakov @mohitban47 @eunsolc @MarieMarneffe !
📢 10 Days Left to apply for the AthNLP - Athens Natural Language Processing Summer School! ✍ Get your applications in before June 15th! https://t.co/2D4hlmXvsn
1
12
24
@pietro_lesci is finishing his PhD soon, and on the job market :) so hire him if you want someone who thinks outside the box!
This lead author (@pietro_lesci) has the most creative NLP papers. He adapts methods typically used for observational studies and the results are always so cool
0
1
6
A string may get 17 times less probability if tokenised as two symbols (e.g., ⟨he, llo⟩) than as one (e.g., ⟨hello⟩)—by an LM trained from scratch in each situation! Our #acl2025nlp paper proposes an observational method to estimate this causal effect! Longer thread soon!
3
24
136
🔥 We teach LLMs to say how confident they are on-the-fly during long-form generation. 🤩No sampling. No slow post-hoc methods. Not limited to short-form QA! ‼️Just output confidence in a single decoding pass. ✅Better calibration! 🚀 20× faster runtime. arXiv:2505.23912 👇
2
22
40
📄 Paper: https://t.co/VMfsSHc2Oj 💻 Code and annotated data: https://t.co/fZ8YM9lFIi (coming soon) Big thank you to my collaborators @vlachos_nlp, @michael_sejr, @nedjmaou!
github.com
Code and data for our paper on automatically uncovering research framings in NLP papers. - ericchamoun/NLP-Framing-Analysis
1
1
3