I have similar observations. But as a reviewer, I have to be honest that I cannot check each claim about previous papers, and these kinds of false references are often considered as minor issues (not really) comparing to novelty or empirical results.

thanks to Leena Mathur and Su Li for helping with collecting robotics videos.
thanks @michaelryan207.bsky.social @williamheld.com @echoshao8899.bsky.social @jyangballin.bsky.social @ellaminzhili.bsky.social @juliakruk.bsky.social @rewang.bsky.social @vidhijain.bsky.social @ybisk.me Dorsa Sadigh

Incredible collab with MohammadHossein Rezaei* (U of A) Yicheng Fu* Phil Cuvin* (U of T) @calebziems.com @yanzhe.bsky.social @diyiyang.bsky.social

Upvote our paper at huggingface.co/papers/2502....
arxiv.org/abs/2502.20490

As always, we open source everything. Even our nicely made website: egonormia.org Please check out the leaderboard, the blog (w/Bibtex support), the code, data, as well as a data viewer.

Visit nnetnav.dev for more examples, code, and data.

The key insight is that LLMs are good at understanding whether a traj is doing something reasonable and that guides efficient exploration and gives accurate labels. Be warned that deploying exploration algorithms in the real world has consequences -- monitor your agents closely.

Our awesome team: @ellaminzhili.bsky.social @williamheld.com @michaelryan207.bsky.social Kunat Pipatanakul, Potsawee Manakul @diyiyang.bsky.social

matplotlib with customization. I can share the code with you

Would really appreciate it if I can be included. I build social intelligence models/agents that can cooperate with humans.

🙋‍♂️