Supryadi

Computer Science PhD student at Tianjin University | A newbie researcher in NLP field

[PluralValueAI2026 @PluralValAI2026]

📢 News from the PlurVA-LLM Shared Task!

✅ Dev & Test Sets have been released.
✅ Leaderboards are now open for submissions.
Registered but didn't receive the datasets? Please check your spam folder first. If you still can't find them, email us at [email protected]

[TJUNLP @TJUNLP]

We are excited to share that TJUNLP has 16 papers accepted in ACL 2026 (9 Main, 7 Findings). Feel free to reach out if you'd like to discuss research or potential collaborations. See you in San Diego! #ACL2026 #NLProc

[PluralValueAI2026 @PluralValAI2026]

📢 2nd Call for Papers!

PlurVA-LLM @aaclmeeting 2026
First Workshop on Pluralistic Value Alignment of LLMs
📍 Hengqin, Zhuhai, China
🗓️ During AACL-IJCNLP 2026, 6–10 Nov 2026
🔗 https://t.co/nXXQ345GTL

[PluralValueAI2026 @PluralValAI2026]

🚀Join the PlurVA-LLM 2026 Shared Task!

We invite participation in “Pluralistic Value Alignment Across China, Indonesia & Sri Lanka” at PlurVA-LLM @aaclmeeting .

🗓️ Deadline: 20 July 2026
🔗Website: https://t.co/Hhy7TgGQRk
📍Register: https://t.co/YHqldkbQTA

[PluralValueAI2026 @PluralValAI2026]

🌐Our Shared Task is officially LIVE!
PlurVA-LLM @aaclmeeting challenges you to build a system that aligns with pluralistic values across:
🇨🇳 China 🇮🇩 Indonesia 🇱🇰 Sri Lanka
🔗Details & registration:
https://t.co/Hhy7TgGQRk
#AACL2026 #SharedTask #PluralistValueAlignment #NLP

[PluralValueAI2026 @PluralValAI2026]

🚨CFP: PlurVA-LLM Workshop @aaclmeeting
We invite work on but not limited to:
• Value alignment
• Multilingual & cross-cultural NLP
• Evaluation & datasets
• AI safety & reasoning
📅Deadline: Aug 15, 2026
🔗 https://t.co/nXXQ345GTL

[PluralValueAI2026 @PluralValAI2026]

Excited to introduce PlurVA-LLM: the 1st Workshop on Pluralistic Value Alignment of LLMs.

📍 Hengqin, China
📅 November, 2026
🔗 https://t.co/qkCa3B8OQR

Call for Papers coming soon! @aaclmeeting #PlurVALLM #PluralistValues #AIAlignment #2026

[Alex Cui @alexcdot]

Okay so, we just found that over 50 papers published at @Neurips 2025 have AI hallucinations

I don't think people realize how bad the slop is right now

It's not just that researchers from @GoogleDeepMind, @Meta, @MIT, @Cambridge_Uni are using AI - they allowed LLMs to generate hallucinations in their papers and didn't notice at all.

It's insane that these made it through peer review👇

[Omar Sanseviero 🇫🇷 @RAISE Paris @osanseviero]

Introducing T5Gemma 2, the next generation of encoder-decoder models 🚀

Built on top of Gemma 3, we were able to build compact models at sizes of 270m-270m, 1B-1B, and 4B-4B sizes.

While most models today are decoder-only, T5Gemma 2 is the first (I'm aware of) multimodal, long-context, and heavily multilingual (140 languages) encoder-decoder model out there.

We hope this model enables the model research community as well as the community of devs ready to explore with new architectures.

Blog: https://t.co/12ScxYcjxa
Models: https://t.co/D38wNFo5Bc
Paper: https://t.co/2rypSQ7Bf6

[Omar Sanseviero 🇫🇷 @RAISE Paris @osanseviero]

Introducing Gemma Scope 2

🤗Largest open release of interpretability tools (over 1 trillion parameters trained!)
🔬Works as a microscope to analyze all Gemma 3 models' internal activations
🗣️Advanced tools for analyzing chat behaviors

[Negar Foroutan @negarforoutan]

1/ 🌍 How does mixing data from hundreds of languages affect LLM training?
In our new paper "Revisiting Multilingual Data Mixtures in Language Model Pretraining" we revisit core assumptions about multilinguality using 1.1B-3B models trained on up to 400 languages.
🧵👇

[TJUNLP @TJUNLP]

How will AI reshape the future of scientific discovery and innovation? We proudly to share our new survey paper: “Towards AI-Driven Science and Innovation: Paradigms, Practices, Challenges, and Horizons”.
📖: https://t.co/YE0jyc1ayD
🌐: https://t.co/yZl5n8TMTV

[TJUNLP @TJUNLP]

TJUNLP is heading to EMNLP 2025 ✈️
We'll be presenting our latest papers there and looking forward to great discussions!
#EMNLP2025 #NLProc

[Yong Zheng-Xin @yong_zhengxin]

🧵 Multilingual safety training/eval is now standard practice, but a critical question remains: Is multilingual safety actually solved?

Our new survey with @Cohere_Labs answers this and dives deep into:
- Language gap in safety research
- Future priority areas

Thread 👇

[Tiago Pimentel @tpimentelms]

If you're finishing your camera-ready for ACL (#acl2025nlp) or ICML (#icml2025 ) and want to cite co-first authors more fairly, I just made a simple fix to do this! Just add $^*$ to the authors' names in your bibtex, and the citations should change :)

https://t.co/pBaLpx0Ykl

[Jirui Qi ✈️ ICML 🇰🇷 @Jirui_Qi]

[1/]💡New Paper
Large reasoning models (LRMs) are strong in English — but how well do they reason in your language?

Our latest work uncovers their limitation and a clear trade-off:
Controlling Thinking Trace Language Comes at the Cost of Accuracy

📄Link: https://t.co/SwCR6CYpdA

[Liwei Jiang @liweijianglw]

Can language models reason about 𝕚𝕟𝕕𝕚𝕧𝕚𝕕𝕦𝕒𝕝𝕚𝕤𝕥𝕚𝕔 𝕙𝕦𝕞𝕒𝕟 𝕧𝕒𝕝𝕦𝕖𝕤 𝕒𝕟𝕕 𝕡𝕣𝕖𝕗𝕖𝕣𝕖𝕟𝕔𝕖𝕤? is accepted to #ACL2025 main conference! See you in Vienna! (https://t.co/4yHWCzmXAi)

[Yanai Elazar @yanaiela]

💡 New ICLR paper! 💡
"On Linear Representations and Pretraining Data Frequency in Language Models":

We provide an explanation for when & why linear representations form in large (or small) language models.

Led by @jack_merullo_ , w/ @nlpnoah & @sarahwiegreffe

[Rheeya Uppaal @RUppaal]

@iclr_conf paper alert! The de facto way to align a model through tuning-based methods like DPO is powerful, yet expensive and prone to jailbreaking. Emerging work on model editing aims to address this, and yet the two approaches are largely siloed. Can we somehow connect them?🧐

[Bryan Wilie @bryanwilie92]

1/ 🚀 Excited to share our new research on interlingual representations in LLMs! We explored whether LLMs truly develop unified interlingual constructs or if their cross-lingual alignments are inconsistent.