ntunlpsg

[ntunlpsg @ntunlpsg]

Our analysis paper on the middle curse is accepted to #ACL2024 !

[Mathieu Ravaut @MatRavox]

Checkout our analysis paper, to be featured in ACL main conference:
https://t.co/k8H9AMN9s3
We investigate the "middle-curse" exposed by the LITM paper for the specific task of abstractive summarizarion. In brief: the middle curse is very much an issue ! w @JotyShafiq @AixinSG

[ntunlpsg @ntunlpsg]

Check out our new work: LLMSanitize, A comprehensive survey and a simple-to-use library for detecting contaminations!

[Caiming Xiong @CaimingXiong]

Are your LLMs highly accurate, or simply contaminated?
As the race to build the best LLM intensifies, clean evaluation is becoming more important than ever, yet contaminated LLMs and benchmarks obfuscate the real performance of models.
Checkout our new work (comprehensive survey + library) at NTU-NLP lab & Salesforce Research on the critical issue of contamination detection in LLMs, cc @ntunlp @MatRavox @D_Boss001 @HailinChen3 @XingxuanLi @RuochenZhao3 @FangkaiJiao @qcwntu @CaimingXiong @JotyShafiq
Paper:
https://t.co/16jXQ1hNXk
Library:
https://t.co/vNgXrto1XJ

[Li Xingxuan @XingxuanLi]

We provide more analysis on:
(1) Single vs. multiple knowledge domains and sources
(2) Parallel vs. dynamic knowledge adapting
(3) Evaluating factuality improvement of the rationales

Please refer to our paper for details!
Paper link: https://t.co/GMPT4Wgnr7
🧵(9/9)

[Enamul Hoque Prince @Enamul_Hoque]

We are presenting a tutorial on NLP+Vis: NLP Meets Visualization @emnlp this morning at 9am local time. Explore more at https://t.co/GL775rUssl. #EMNLP2023
@JotyShafiq @jesse_vig

[ntunlpsg @ntunlpsg]

Check out our survey on open-source LLMs that catch up with #ChatGPT !

[Caiming Xiong @CaimingXiong]

It has been exactly one year since the release of #ChatGPT. How far are open-source LLMs? We provide an exhaustive review of open-source LLMs that claim to catch up with or surpass ChatGPT in various capabilities.

paper🔗: https://t.co/ARg0Jt0LjX

[M Saiful Bari (MARUF) @sbmaruf]

xCodeEval: An 🔥Execution-based 🔥Large Scale Multilingual Multitask Benchmark for Code Understanding, Generation, Translation, and Retrieval
🏆 16.5B Tok
🏆 25M Doc
🥳 11 prog. lang.
⏬ Debunking memorization of ChatGPT💥
⏬ Reasoning Spectrum of LLMs 🧐
⏬ChatGPT analysis.👇

[ntunlpsg @ntunlpsg]

A new Execution based evaluation benchmark for Programming Languages. Lots of analysis included.

[M Saiful Bari (MARUF) @sbmaruf]

xCodeEval: An 🔥Execution-based 🔥Large Scale Multilingual Multitask Benchmark for Code Understanding, Generation, Translation, and Retrieval
🏆 16.5B Tok
🏆 25M Doc
🥳 11 prog. lang.
⏬ Debunking memorization of ChatGPT💥
⏬ Reasoning Spectrum of LLMs 🧐
⏬ChatGPT analysis.👇

[ntunlpsg @ntunlpsg]

New Evaluation of ChatGPT with Human in a loop. Lots of new observations!

[M Saiful Bari (MARUF) @sbmaruf]

🔥A Systematic Study and Comprehensive Evaluation of ChatGPT on Benchmark Datasets🔥
Yet another ChatGPT evaluation !!!! What's new ?
This time not automatic, human in a loop.
https://t.co/YeIs8cfiSa
Our ACL'23 paper covers FULL eval on benchmarks that actually MATTERS !!!

[Ruochen Zhao @RuochenZhao3]

❌📖✅
How can we make LLMs output more correct answers? When humans are uncertain, we search for external references before answering. Now, we integrate exactly the same process into LLMs. Our Verify-and-Edit surpasses previous methods in QA tasks.
Paper: https://t.co/AR7UIlSta1

[Mathieu Ravaut @MatRavox]

Not happy about the abstractive summaries that you are generating from the language model?
Have a try at sampling many and re-ranking them! 👇
https://t.co/FbOTMXnIVq

[Mathieu Ravaut @MatRavox]

Check out our new paper ! 🔥
To be featured at EMNLP 2022
#emnlp2022 #NLProc

[ntunlpsg @ntunlpsg]

(3/3) Contrastive Learning" by Qian Li
"Data Selection Curriculum for Neural Machine Translation" by @mtaasnim
"BotSIM: An End-to-End Bot Simulation Framework for Commercial Task-Oriented Dialog Systems" by Guangsen Wang and @samsontmr

[ntunlpsg @ntunlpsg]

(2/3) "AdaPT: A Set of Guidelines for Hyperbolic Multimodal Multilingual NLP" by @ramitsawhney, @megh_1211 et al.
"Enhancing Multilingual Language Model with Massive Multilingual Knowledge Triples" by Linlin Liu et al.
"Alleviating Sparsity of Open Knowledge Graphs with Ternary

[ntunlpsg @ntunlpsg]

(1/3) We have 8 papers accepted to #emnlp2022!
"Towards Summary Candidates Fusion" by @MatRavox
"Learning Label Modular Prompts for Text Classification in the Wild" by @HailinChen3
"OpenCQA: Open-ended Question Answering with Charts" by Shankar Kantharaj and @dxlong2000 et al.

[ntunlpsg @ntunlpsg]

Two of our works are featured at #NeurIPS2022: "SleepQA: A Health Coaching Dataset on Sleep for Extractive Question Answering" with @megh_1211, and "Refining Low-Resource Unsupervised Translation by Language Disentanglement of Multilingual Translation Model" by Nguyen Xuan Phi.

[ntunlpsg @ntunlpsg]

Our paper "Enhancing Multilingual Language Model with Massive Multilingual Knowledge Triples" was accepted by #emnlp2022! We present novel knowledge-based multilingual language models (KMLMs) trained directly on knowledge triples. https://t.co/JORcYd7rtK

[Prathyusha Jwalapuram @jwala_94]

I'll be presenting our work on "Rethinking Self-Supervision Objectives for Generalizable Coherence Modeling" at #ACL2022 this afternoon (2.30pm) in the Discourse and Pragmatics session. Do drop by! Full paper: https://t.co/gMicpEskle

[ntunlpsg @ntunlpsg]

@SRoyLee @prajdabre1 @iclr_conf @aclmeeting Hello! Thanks for the interest! Chart-to-text is going through a few final changes, the camera-ready version will be out next week. We'll put out an arXiv link next week and tag you in the comments :)

[ntunlpsg @ntunlpsg]

Our group had 2 papers accepted to ICLR @iclr_conf and 6 papers accepted to ACL @aclmeeting this year, including SummaReranker, Chart-to-Text, LEPT5.

[ntunlpsg @ntunlpsg]

'GlobalWoZ: Globalizing MultiWoZ to Develop Multilingual Task-Oriented Dialogue Systems' has been accepted to #ACL2022 ! It introduced a novel data curation method that generates GlobalWoZ --- a large-scale multilingual ToD dataset. Paper at: https://t.co/XFSZA6pv6U

[ntunlpsg @ntunlpsg]

Our paper "Rethinking Self-Supervision Objectives for Generalizable Coherence Modeling" by @jwala_94 has been accepted at #ACL2022! It proposes a SOTA coherence model that generalizes beyond the permuted document task to downstream applications. Paper: https://t.co/3dvakRrnWP