A lot of students I know, who deserve to attend EMNLP 2026, will miss the chance due to high registration fees and accommodation costs. Isn't there any way to make some things cheaper? I know that D&I do a lot of good work, but can't big sponsors like FAANG also help?

We take a closer look at prompt optimization problem & develop UniPrompt for learning LM prompt for a task from scratch. We argue when greedy opt. methods could help in this discrete & challenging problem. Gurusha Juneja, Amit Sharma, Hua Li & Jian Jiao. arxiv.org/pdf/2406.10504

UNIPROMPT 现有的 Prompt 优化方法往往生成简短的 Prompt，无法充分捕捉任务的复杂性，人工编写 Prompt 时会考虑任务的多个方面，但现有算法难以做到这一点。 来自 Microsoft Research 和 Microsoft Bing Ads 的研究者们推出了 UNIPROMPT，它模仿人类编写 Prompt 的过程，通过两个阶段生成复杂的

How can you scale up robot learning? We introduce HOP: Extract hand-object trajectories from in-the-wild videos. Train a next-token-prediction model to get a manipulation prior. Adapt to your task using BC or RL. Checkout our webpage: bit.ly/47tyeDv More details in 🧵

It's a blockbuster performance by our lab LCS2 Lab at #EMNLP2024. 8 out of 8 papers have been accepted. Our gang is coming to Miami. 👊 EMNLP 2025 IIT Delhi

First year PhD Gurusha Juneja and Rutgers Prof/Alumna Sharon Levy presenting their work at #EMNLP2024

So no AGI in 2024 then? 🤔

You have no idea how badly I wanted to shout this info out for the past 2 months. Finally, I can talk about it. Yes, AACL is in IIT Bombay. Mumbai is my domain and I, along with the local team, will make your time there memorable. Please stay tuned and consider submitting your

Finally, a conference in India! Can't explain how happy it makes me to hear this news.

Congratulations Xinyi Wang! Best wishes for your journey at Princeton and SUNY Buffalo.

Just read "What Really Matters in Matrix Whitening Optimizers?" by Kevin Frans. I genuinely find this paper clear and insightful. It deepened my understanding of pre-conditioners and the geometry behind more optimizers. ICLR 2026 scores (2,2,2,4) don’t reflect the quality here.

Woke up to a bed tea today - All reviews/scores are being rolled back to pre-discussion states. No more reviewer discussion. ICLR 2026 Open Review Month end came with a little too much masala.

At #NeurIPS2025 this week! Let’s talk about: 💻 LLM Agents 🧠 Multi-Agent Learning 🔬 AI for Scientific Discovery 🌟 Curiosity-driven Exploration 🤖 Embodied AI …or just California sunshine in December🌞 Would love to meet folks — DM or stop me in the halls!

Alyosha’s energy is contagious — this is what research passion looks like. Presenting his student’s work Anand Bhattad Konpat Ta Preechakul at #NeurIPS2025

Crossed paths with Prof. Pratyusha Sharma ✈️ NeurIPS near her poster “Lora vs Full Fine-tuning” at #NeurIPS2025. Her work offers fundamental insights for the field and has strong scientific rigour. Had a discussions on research directions and long-term vision. Was nice to briefly connect.

Does anyone else hate that OpenAI over optimized for skimmability? ChatGPT's lists in every response are annoying.

Human perception is inherently situated – we understand the world relative to our own body, viewpoint, and motion. To deploy multimodal foundation models in embodied settings, we ask: “Can these models reason in the same observer-centric way?” We study this through SAW-Bench: