UIUC NLP

@uiuc_nlp

Natural Language Processing research group at The University of Illinois Urbana-Champaign @IllinoisCS @UofIllinois

Joined July 2019

586Posts 1KFollowers 140Following

You might like

@hengjinlp

@jefffhj

@kaiwei_chang

@xiangrenNLP

@uwnlp

@hhexiy

@WeijiaShi2

@VioletNPeng

@HannaHajishirzi

@ManlingLi_

@lileics

@unc_ai_group

@hhsun1

@ysu_nlp

@jhuclsp

UIUC NLP reposted

Alexi Gladstone

@AlexiGlad

4 h

What if your policy could reason and think dynamically, especially about uncertainty, enabling better real-world behavior? ⚡️Introducing EBT-Policy, an instantiation of Energy-Based Transformers for Policies! TLDR: - EBT-Policy broadly outperforms Diffusion Policy in both…

AlexiGlad's tweet image. What if your policy could reason and think dynamically, especially about uncertainty, enabling better real-world behavior?

⚡️Introducing EBT-Policy, an instantiation of Energy-Based Transformers for Policies!
TLDR:
- EBT-Policy broadly outperforms Diffusion Policy in both…

UIUC NLP reposted

Heng Ji

@hengjinlp

Nov 2

Many of my students cannot attend EMNLP in person due to visa problems, but the super rising star Cheng Qian @qiancheng1231 will be there presenting multiple papers. Please drop by our posters and talk to him!

UIUC NLP reposted

Diyi Yang

@Diyi_Yang

Oct 28

@ZhiruoW 's research compares AI agents vs humans across real work tasks (data analysis, engineering, design, writing). Key findings: 👉Agents are 88% faster & 90-96% cheaper 👉BUT produce lower quality work, often fabricate data to mask limitations 👉Agents code everything,…

Diyi_Yang's tweet image. @ZhiruoW 's research compares AI agents vs humans across real work tasks (data analysis, engineering, design, writing). Key findings:

👉Agents are 88% faster &amp; 90-96% cheaper
👉BUT produce lower quality work, often fabricate data to mask limitations
👉Agents code everything,…

Zora Wang

@ZhiruoW

Oct 28

Agents are joining us at work -- coding, writing, design. But how do they actually work, especially compared to humans? Their workflows tell a different story: They code everything, slow down human flows, and deliver low-quality work fast. Yet when teamed with humans, they shine…

UIUC NLP reposted

Stanford NLP Group

@stanfordnlp

Oct 25

Today, we’re overjoyed to have a 25th Anniversary Reunion of @stanfordnlp. So happy to see so many of our former students back at @Stanford. And thanks to @StanfordHAI for the venue!

stanfordnlp's tweet image. Today, we’re overjoyed to have a 25th Anniversary Reunion of @stanfordnlp.

So happy to see so many of our former students back at @Stanford.

And thanks to @StanfordHAI for the venue!

UIUC NLP reposted

Haoyi Qiu

@HaoyiQiu

Oct 28

🤖💬AI agents can be easily persuaded (like Anthropic’s Claudius often giving discounts). 🤔Previous study on persuasion has been exclusively on text-only modality. We wonder: are AI agents more susceptible when presented with multimodal content? Introducing MMPersuade, a…

HaoyiQiu's tweet image. 🤖💬AI agents can be easily persuaded (like Anthropic’s Claudius often giving discounts).

🤔Previous study on persuasion has been exclusively on text-only modality. We wonder: are AI agents more susceptible when presented with multimodal content?

Introducing MMPersuade, a…

UIUC NLP reposted

Manling Li

@ManlingLi_

Oct 17

World Model Reasoning for VLM Agents (NeurIPS 2025, Score 5544) We release VAGEN to teach VLMs to build internal world models via visual state reasoning: - StateEstimation: what is the current state? - TransitionModeling: what is next? MDP → POMDP shift to handle the partial…

UIUC NLP reposted

Zhenhailong Wang

@zhenhailongW

Oct 14

Multimodal conversational agents struggle to follow complex policies, which also impose a fixed computational cost. We ask: 👉 How can we achieve stronger policy-following behavior without having to include policies in-context? 🌐: mikewangwzhl.github.io/TriMPI/ 🧵1/3

zhenhailongW's tweet image. Multimodal conversational agents struggle to follow complex policies, which also impose a fixed computational cost.
We ask:
👉 How can we achieve stronger policy-following behavior without having to include policies in-context?
🌐: mikewangwzhl.github.io/TriMPI/ 🧵1/3

UIUC NLP reposted

Yiling Lou

@yiling__LOU

Oct 6

Thrilled to announce that I'll be joining UIUC CS @siebelschool as an Assistant Professor in Spring 2026! 📢 I’m looking for Fall '26 PhD students who are interested in the intersection of Software Engineering and AI, especially in LLM4Code and Code Agents. Please drop me an…

UIUC NLP reposted

Shizhe Diao

@shizhediao

Oct 6

🚀 Introducing BroRL: Scaling Reinforcement Learning via Broadened Exploration When step-scaling hits a plateau, scale rollouts, not steps. BroRL takes reinforcement learning beyond saturation—reviving stalled models by expanding exploration with large-N rollouts. 👇 (1/n)

shizhediao's tweet image. 🚀 Introducing BroRL: Scaling Reinforcement Learning via Broadened Exploration

When step-scaling hits a plateau, scale rollouts, not steps.
BroRL takes reinforcement learning beyond saturation—reviving stalled models by expanding exploration with large-N rollouts.
👇 (1/n)

UIUC NLP reposted

Prashant Jayannavar

@p_jayannavar

Oct 5

🚨 New preprint out! 🚨 In "BAP v2: An Enhanced Task Framework for Instruction Following in Minecraft Dialogues", we work towards a core AI challenge: how can agents follow complex, conversational instructions in a dynamic 3D world? To this end, we introduce an enhanced task…

p_jayannavar's tweet image. 🚨 New preprint out! 🚨 In "BAP v2: An Enhanced Task Framework for Instruction Following in Minecraft Dialogues", we work towards a core AI challenge: how can agents follow complex, conversational instructions in a dynamic 3D world? To this end, we introduce an enhanced task…

UIUC NLP reposted

Se-woong (Sam) Lee

@sewoong_sam_lee

Oct 4

🚨 New paper alert at COLM 2025! 🚨 An interesting open problem for those into Sparse Autoencoders (SAEs): "Top-K activation constrains L0 (the number of non-zeros), but how do we obtain E[L0]?" This was even the very first limitation noted in @nabla_theta’s recent paper. (1/8)

sewoong_sam_lee's tweet image. 🚨 New paper alert at COLM 2025! 🚨
An interesting open problem for those into Sparse Autoencoders (SAEs):
"Top-K activation constrains L0 (the number of non-zeros), but how do we obtain E[L0]?"

This was even the very first limitation noted in @nabla_theta’s recent paper. (1/8)

UIUC NLP reposted

EMNLP 2025

@emnlpmeeting

Oct 1

Please note that #EMNLP2025 volunteer notifications have been sent. If you haven’t received yours, please check your spam folder or contact the chairs at [email protected] as some email addresses were entered incorrectly in the form

UIUC NLP reposted

Bowen Jin

@BowenJin13

Oct 1

Very excited to see Tinker released by @thinkymachines! Even more thrilled that Search-R1 is featured as the tool-use application in Tinker’s recipe 👇 🔗 github.com/thinking-machi… When we first built Search-R1, we opened up everything—data, recipes, models, code, logs—and kept…

Thinking Machines

@thinkymachines

Oct 1

Introducing Tinker: a flexible API for fine-tuning language models. Write training loops in Python on your laptop; we'll run them on distributed GPUs. Private beta starts today. We can't wait to see what researchers and developers build with cutting-edge open models!…

thinkymachines's tweet image. Introducing Tinker: a flexible API for fine-tuning language models.

Write training loops in Python on your laptop; we'll run them on distributed GPUs.

Private beta starts today. We can't wait to see what researchers and developers build with cutting-edge open models!…

UIUC NLP reposted

Alexi Gladstone

@AlexiGlad

Sep 30

*Human-Like* Creativity is perhaps the most out-of-reach task for modern LLMs I'm super excited to share our new work evaluating LLMs with a creativity framework! We develop a synthetic creativity task to measure LLMs' capabilities in generating novel, creative, combinations,…

Samuel Schapiro

@samschapiro

Sep 30

1/N Large language models (LLMs) have been widely adopted for closed-ended tasks like reasoning, but can they truly be creative? 📚 Excited to announce our new work — Combinatorial Creativity: A New Frontier in Generalization Abilities. 📝 Paper arxiv.org/abs/2509.21043

samschapiro's tweet image. 1/N Large language models (LLMs) have been widely adopted for closed-ended tasks like reasoning, but can they truly be creative?

📚 Excited to announce our new work — Combinatorial Creativity: A New Frontier in Generalization Abilities.

📝 Paper arxiv.org/abs/2509.21043

UIUC NLP reposted

Kung-Hsiang Steeve Huang

@steeve__huang

Oct 2

Ever felt like your GUI agents are dragging their feet? 🧐The culprit? Crunching through endless streams of screenshots, especially in those marathon long-horizon tasks. Thrilled to unveil ⭐️ GUI-KV ⭐️— our plug-and-play powerhouse that taps into the spatial saliency within…

steeve__huang's tweet image. Ever felt like your GUI agents are dragging their feet? 🧐The culprit? Crunching through endless streams of screenshots, especially in those marathon long-horizon tasks.

Thrilled to unveil ⭐️ GUI-KV ⭐️— our plug-and-play powerhouse that taps into the spatial saliency within…

UIUC NLP reposted

Cheng Qian

@qiancheng1231

Sep 26

🚀 Introducing UserRL: a new framework to train agents that truly assist users through proactive interaction, not just chase static benchmarking scores. 📄 Paper: arxiv.org/pdf/2509.19736 💻 Code: github.com/SalesforceAIRe…

qiancheng1231's tweet image. 🚀 Introducing UserRL: a new framework to train agents that truly assist users through proactive interaction, not just chase static benchmarking scores.

📄 Paper: arxiv.org/pdf/2509.19736
💻 Code: github.com/SalesforceAIRe…

UIUC NLP reposted

UIUC TIMAN Group

@TIMANUIUC

Aug 27

🧠 Get $20 to watch videos + share what sparks your curiosity! TIMAN Lab @ UIUC is studying how people seek info while watching content. 🖥️ 30–60 min Zoom session ✅ 18+, fluent in English 💸 $20 reward Sign up: docs.google.com/forms/d/1OcnrM… #PaidStudy #AI

UIUC NLP reposted

Heng Ji

@hengjinlp

Sep 19

Accepted as NeurIPS2025 Spotlight! Existing large multimodal models (LMMs) have very poor visual understanding and reasoning over part-level attributes and affordances (only 5.9% gIoU). We developed novel part-centric LMMs to address these challenges arxiv.org/pdf/2505.20759

hengjinlp's tweet image. Accepted as NeurIPS2025 Spotlight! Existing large multimodal models (LMMs) have very poor visual understanding and reasoning over part-level attributes and affordances (only 5.9% gIoU). We developed novel part-centric LMMs to address these challenges arxiv.org/pdf/2505.20759

UIUC NLP reposted

Siebel School of Computing and Data Science

@siebelschool

Sep 19

Exciting growth ahead! 🚀 Eighteen new faculty members are joining the Siebel School of Computing and Data Science at Grainger Engineering, bringing world-class expertise in research and education. ▶️bit.ly/3VJ281D

siebelschool's tweet image. Exciting growth ahead! 🚀 Eighteen new faculty members are joining the Siebel School of Computing and Data Science at Grainger Engineering, bringing world-class expertise in research and education.

▶️bit.ly/3VJ281D

UIUC NLP reposted

Cheng Qian

@qiancheng1231

Sep 19

Our paper is accepted to NeurIPS 2025 for presentation! Please check my post below for paper details! 🚀🚀🚀

Cheng Qian

@qiancheng1231

Apr 22

🚀 ToolRL unlocks LLMs' true tool mastery! The secret? Smart rewards > more data. 📖 Introducing newest paper: ToolRL: Reward is all Tool Learning Needs Paper Link: arxiv.org/pdf/2504.13958 Github Link: github.com/qiancheng0/Too…

qiancheng1231's tweet image. 🚀 ToolRL unlocks LLMs' true tool mastery! The secret? Smart rewards &gt; more data.

📖 Introducing newest paper:
ToolRL: Reward is all Tool Learning Needs

Paper Link: arxiv.org/pdf/2504.13958
Github Link: github.com/qiancheng0/Too…

Heng Ji

@hengjinlp

Yu Su

@ysu_nlp

Kyunghyun Cho

@kchonyc

Greg Durrett

@gregd_nlp

Bill Yuchen Lin

@billyuchenlin

Delip Rao e/σ

@deliprao

Kuan-Hao Huang

@kuanhaoh_

Jie Huang

@jefffhj

Pan Lu

@lupantech

Jessy Li

@jessyjli

Soumya Sanyal

@ssanyal8

$sunjiao123sun_'s profile picture. Senior RS @ GoogleDeepMind, improving Gemini Coding \n\n NLP PhD @ USC, Amazon ML Fellow \n\n ex-{Google Brain, Alexa AI} nlper, IIIS Tsinghua-Ren$

Jiao Sun

@sunjiao123sun_

Preslav Nakov

@preslav_nakov

Kenneth Huang

@windx0303

Revanth Gangi Reddy

@gangi_official

Shaily

@shaily99

Mohit Bansal

@mohitban47

Lakshya

@Lakshya604

Haoran Wang

@hwang2226

Aniket Vashishtha

@AniketVashisht8

paul

@paul_okewunmi

jai sharma

@jaishrm07

Kunal Samanta

@frankensteinllm

λ

@lambda_functor

Priyanshu

@Priyans62580298

Nicole

@Irxirmo89322

yydxlv

@yydxlv

Nadine

@halbrook5863683

Zhihui Yang

@Zhihui_Yang_

Eric Connelly

@ericmconnelly

28X12x24

@xvvv19zx0

schlotti

@gyybvb

Bo Shui

@brian_cyan1

Nalin Wadhwa

@nalin_wadhwa

Jimeng Shi

@jimengshi008

Sameer Deshmukh

@v0dro

Lei Wang

@MajorTom414

Yi-Jyun Sun

@jyun_sun

Tristan

@tristan4ai

Ritwik

@heyitsrits

SAI RAM

@rsai31385_ram

Arturo

@arturoComputes

Subramanyam Sahoo

@iamwsubramanyam

sdp

@gucciga74321588

Yiling Lou

@yiling__LOU

keshav baniya

@kingership321

Neahas

@Neahas79662

Pragathi Y P

@PragathiYP23

hobart

@hobart24928224

Nathan

@Nathanomics2169

Zhaoyi ZHANG

@ZhaoyiZHANG66

Cheng Wang

@WangCheng_0116

Justin

@Justin_6657

Shubhankar Tripathy

@shubhtrips

Min

@aooimin

Kunlun Zhu

@Kunlun_Zhu

Chris Kong

@Chris61643492

The_Scholar📚🖊️

@Sabiu_Idreees

Wenhao Yu

@wyu_nd

Shyamal Narang

@NarangShyamal

Bellina Chu

@brrbaow

Junker

@junk1r

smile

@Smilex_P

Rutva Pandya

@RutvaPandya8

RockyParadox

@RockyParadox44

AG

@pon55272377

Katherine Tieu

@kthrn_uiuc

(((ل()(ل() 'yoav))))👾

@yoavgo

Graham Neubig

@gneubig

Heng Ji

@hengjinlp

Sasha Rush

@srush_nlp

William Wang

@WilliamWangNLP

Kyunghyun Cho

@kchonyc

Greg Durrett

@gregd_nlp

Rada Mihalcea

@radamihalcea

Kuan-Hao Huang

@kuanhaoh_

Jie Huang

@jefffhj

Christopher Manning

@chrmanning

Jessy Li

@jessyjli

Princeton NLP Group

@princeton_nlp

$sunjiao123sun_'s profile picture. Senior RS @ GoogleDeepMind, improving Gemini Coding \n\n NLP PhD @ USC, Amazon ML Fellow \n\n ex-{Google Brain, Alexa AI} nlper, IIIS Tsinghua-Ren$