First time coming to the sea.

Introducing 🔥SFR-RAG 🔥, an LLM specialized in RAG use cases, optimized to fully leverage contextual content for accurate and faithful response generation. We also introduce ContextualBench - A compilation of 7 popular contextual QA tasks measured in different settings. On

big ups to Albert for an inspiring First Look talk at #WMIF2024. pls check it out! youtube.com/watch?v=axxTJM…… Mass General Cancer Center Damon Runyon Cancer Research Foundation MGH Martinos Center

Can your LLM Stay Faithful to Context, Even If "The Moon 🌕 is Made of Marshmallows 🍡"? We Introduce FaithEval, a new and comprehensive benchmark dedicated to evaluating contextual faithfulness for LLMs with 4.9K high-quality question-context pairs across 3 challenging tasks:

.MassGeneral News and Harvard Medical School school Investigator Gary Ruvkun, PhD, has won the 2024 The Nobel Prize in Physiology or Medicine. He is honored for his discovery of small regulatory RNAs. Congratulations! Learn more: massgeneralbrigham.org/en/about/newsr…

A junior from BITS Pilani, Goa shared today's mid-sem exam paper for their 'Generative AI' course. Coincidentally, one of the questions (Q4) was on Hopfield Networks—right as John Hopfield was awarded the Nobel Prize in Physics for it today. So cool!

I'm on the job market! Please reach out if you are looking to hire someone to work on - RLHF - Efficiency - MoE/Modular models - Synthetic Data - Test time compute - other phases of pre/post-training. If you are not hiring then I would appreciate a retweet! More details👇

no AI here, just the coolest paper i've seen in a while

This deep tech startup is solving air pollution in India. They make affordable, powerful, low maintenance (filterless), high-volume air purification systems for both factories (MK II) and homes (Hive). It's called Praan, and here's their incredible story: 1/5

Grateful to MIT CSAIL Alliances for having me on their podcast! It was a joy sharing how our recent work trains machines to self-adapt to new tasks and scenarios! Paper: lnkd.in/g2g-RcDb Podcast: bit.ly/40IVZ7v

🚨 New Research Alert! 🚨 Happy to share my latest work - MedHallu: A Comprehensive Benchmark for Detecting Medical Hallucinations in Large Language Models 📄Arxiv - arxiv.org/abs/2502.14302 🤗HuggingFace Dataset - huggingface.co/datasets/UTAus… 🌐Website - medhallu.github.io

how it feels to do anything with claude

Autists had a great run, the AI future belongs to ADHD

🚨New preprint🚨 I’m super excited to share our work: An Empirical Study on Strong-Weak Model Collaboration for Repo-level Code Generation 📜: arxiv.org/abs/2505.20182 w/ Atharva Naik , Yiqing Xie and Carolyn Rose 🧵

most foundational concept in deep learning that no one understands is probably the Neural Tangent Kernel (NTK) this line of work studies neural networks of *infinite width*, which explain a lot about normal finite-width NNs and there is exactly one Very Good blog post on them:

Excited to share that our paper, “MedHallu: A Comprehensive Benchmark for Detecting Medical Hallucinations in Large Language Models,” was accepted to EMNLP 2025 (Main)! Huge thanks to my co-authors Jiawei Xu, Junyuan "Jason" Hong, Atlas Wang, Tianlong Chen, Kaidi Xu, and Ying Ding!

Meet SFR-DeepResearch (SFR-DR) 🤖: our RL-trained autonomous agents that can reason, search, and code their way through deep research tasks. 🚀SFR-DR-20B achieves 28.7% on Humanity's Last Exam (text-only) using only web search 🔍, browsing 🌐, and Python interpreter 🐍,

📊Excited to introduce our Hard2Verify from Salesforce AI Research , the benchmark for measuring the ability of verifiers to provide step-level correctness labels for model-generated responses to largely open-ended, frontier-level math problems📚 The current paradigm of RLVR requires

One of the key challenges for building web-based “deep research” agents is to construct sufficiently difficult long-horizon agentic data. At Salesforce AI Research, We introduce ProgSearch, a controlled data synthesis pipeline that builds tasks of increasing complexity until a frontier