📢Is current “human-AI alignment” research clarified and comprehensive? 🤔 We systematically reviewed 400+ papers across HCI, NLP, and ML to develop a framework for 👫<>🤖"Bidirectional Human-AI Alignment", encompassing the dual paths of “Aligning AI to Human” and “Aligning Human

2/ 💎【Bidirectional Human-AI Alignment Framework】 We introduce our “Bidirectional Human-AI Alignment” framework developed from the systematic review. 🔸A🔸 “Align AI to Human” focuses on mechanisms ensuring AI systems’ objectives match those of humans’. 🔸B🔸 “Align Humans to

I had a lot of fun talking with Lex Fridman about a wide range of topics on his podcast, alongside Dario and Chris. Hope it's interesting to others!

Can VLMs build Spatial Mental Models like humans? Reasoning from limited views? Reasoning from partial observations? Reasoning about unseen objects behind furniture / beyond current view? Check out MindCube! 🌐mll-lab-nu.github.io/mind-cube/ 📰arxiv.org/pdf/2506.21458

Slides for my lecture “LLM Reasoning” at Stanford CS 25: dennyzhou.github.io/LLM-Reasoning-… Key points: 1. Reasoning in LLMs simply means generating a sequence of intermediate tokens before producing the final answer. Whether this resembles human reasoning is irrelevant. The crucial

📢 Excited to share our new paper, where we introduce, ✨GrAInS✨, an inference-time steering approach for LLMs and VLMs via token attribution. Some highlights: ➡️GrAIns leverages contrastive, gradient-based attribution to identify the most influential textual or visual tokens

Just got our Neurips reviews back, and the reviewers nailed every limitation and missing experiment we already knew about (and no, these weren't LLM-generated😅). Here's a thought: authors are often their own harshest critics. What if conferences required us to submit detailed

still waiting for someone to build IMDB for academia🫥 no one should waste 5 years on a badly reviewed movie

We are hiring! Interested in optimizing/scaling RL framework for pretrain scale RL? DM me or apply here: job-boards.greenhouse.io/xai/jobs/47991…

Insane AI news A new paper introduces ASI-ARCH, a fully automated AI research loop that can independently discover superior neural network architectures, outpacing human designed models. Unlike traditional methods limited by human trial and error, ASI-ARCH connects LLM based

Theory: Everything Exists in Latent Space At a certain scale of base model intelligence, latent space contains nearly every possible idea, invention, and technology, along with every combination of those things. Therefore, AGI exists in latent space. AGI isn’t something we

Who is Adam? In this video, meet this guy who has some momentum and his cousin AdamW. youtu.be/1_nujVNUsto

Survery paper with lots of insights on Continual Reinforcement Learning Detailed review of existing works, organizing and analyzing their metrics, tasks, benchmarks, and scenario settings. 🧩 Why the field exists A classic RL agent hones one policy for one environment then

"Heartbreak at Headingley, heartbreak at Lord's, and yet a performance with so much heart in Manchester." As always, amazing commentary by Cricbuzz, and what a match! While it ended in a draw, it was probably the best fighting performance by Team India in recent times.

Agentic Reinforced Policy Optimization

📅 10-Year ToT Award (2015) Thang Luong, Hieu Pham & Christopher D. Manning: “Effective Approaches to Attention-based Neural Machine Translation” EMNLP 2015 🔗 aclanthology.org/D15-1166/ A milestone in neural MT and attention mechanisms. 🔁🧠

Amazing! I have been looking for something like this. Probably the best explanation of Flow Matching. Thanks for sharing!

Pranav agarwal ;)