🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

How can LLMs decide when to rely on tools vs. their internal parametric knowledge? 🤖🚀Excited to share our work SMART! 🧠🛠️ Inspired by human metacognition, we enhance LLM self-awareness to reduce tool overuse while boosting performance. Our experiments show that SMARTAgent

Last week, Emre Can Acikgoz at ConvAI@UIUC @IllinoisCDS released CoALM 🚀 fully open-source Conversational Agentic Language Models with CoALM 8B, CoALM 70B, and CoALM 405B. Excelling in both multi-turn dialogue management & function calling, these models were trained using Oumi

[1/6] Can LLMs out-persuade each other? 🤖🧠💬 Introducing Persuade Me If You Can (PMIYC)—a new framework to evaluate (1) how persuasive LLMs are and (2) how easily they can be persuaded! 🚀 📄Arxiv: arxiv.org/abs/2503.01829 🌐Project Page: beyzabozdag.github.io/PMIYC/

While persuasive models are promising for social good, they can also be misused towards harmful behavior. Recent work by Beyza Bozdag and Shuhaib Mehri aims to assess LLM persuasiveness and susceptibility towards persuasion.

Era of conversational embodied agents is here! Check the latest blogpost by Vardhan Dongre ConvAI@UIUC

Agents like OpenAI Operator can solve complex computer tasks, but what happens when users use them to cause harm, e.g. automate hate speech and spread misinformation? To find out, we introduce SafeArena (safearena.github.io), a benchmark to assess the capabilities of web

LLM alignment doesn't transfer to Web Agents. SafeArena is a simple web environment and testbed to test the safety of agents, built on WebArena. A huge team effort that was highly self-driven 💪 safearena.github.io

Our paper submission deadline is approaching fast! ✍️ Abstract deadline: 21st April Paper deadline 28th April Come and join us in beautiful Avignon, France, to discuss discourse and dialogue 🎉 We invite submissions of original research (long papers, short papers, and demos)

Talking about "DeepSeek-R1 Thoughtology: Let’s <think> about LLM reasoning" Going live at 11am PDT (i.e., 20 mins). Last minute change of plans. You might be able to see live here: youtube.com/watch?v=aO_cTI…

we're launching 🤗 yourbench today, an open source tool for custom benchmarking and synthetic data generation from ANY of your documents. it's a big step towards improving how model evaluations work early access link in replies! (1/8)

Congratulations Sumuk! The core of YourBench is the sophisticated question generation approach for a given document. Check the paper for details. Great collaboration with Hugging Face and ConvAI@UIUC teams.

Try it here: huggingface.co/spaces/yourben… Great work by Sumuk Clémentine Fourrier 🍊 Alina Lozovskaya Gokhan Tur dilek hakkani-tur 🔥

This is an important milestone for enabling LLM-based agents. Reward is all you need for Tool Learning! GRPO achieves significant improvements over base and SFT models for BFCL v3, API-Bank and Bamboogle agentic benchmark tasks. Congratulations Emre Can Acikgoz and Cheng Qian

What do we want from "Conversational Agents" on top of language agents? What is missing in the current Conversational Agent systems? Here is our desideratum with a comprehensive survey of the recent advances in the field. Bonus is a live github collection of new papers organized

Incredibly proud of my students Ada Tur and Gaurav Kamath for winning a SAC award at #NAACL2025 for their work on assessing how LLMs model constituent shifts. Humans have a tendency to move heavier constituents towards the end of the sentence. While LLMs unsurprisingly show

Congratulations to Mila members Ada Tur, Gaurav Kamath and Siva Reddy for their SAC award at #NAACL2025! Check out Ada's talk in Session I: Oral/Poster 6. Paper: arxiv.org/abs/2502.05670

🚀Excited to share our new evaluation paper "TD-Eval: Revisiting Task-Oriented Dialogue Evaluation by Combining Turn-Level Precision with Dialogue-Level Comparisons"! 🔄 🤖⚔️TOD systems have rapidly evolved thanks to LLMs, but traditional metrics remain insufficient in

Thrilled to announce our new survey that explores the exciting possibilities and troubling risks of computational persuasion in the era of LLMs 🤖💬 📄Arxiv: arxiv.org/pdf/2505.07775 💻 GitHub: github.com/beyzabozdag/Pe…

Would models know more about Indian food in Hindi and Turkey’s history in Turkish? Does the language of a question affect an LLM’s answer? ✨Yes!✨ Ishika Agarwal and I are excited to announce our newest preprint in which we explore “Language Specific Knowledge (LSK)”.