Hao Sheng

Very excited about the "First Proof" challenge. I believe novel frontier research is perhaps the most important way to evaluate capabilities of the next generation of AI models. We have run our internal model with limited human supervision on the ten proposed problems. The problems require expertise in their respective domains and are not easy to verify; based on feedback from experts, we believe at least six solutions (2, 4, 5, 6, 9, 10) have a high chance of being correct, and some further ones look promising. We will only publish the solution attempts after midnight (PT), per the authors' guidance - the sha256 hash of the PDF is d74f090af16fc8a19debf4c1fec11c0975be7d612bd5ae43c24ca939cd272b1a . This was a side-sprint executed in a week mostly by querying one of the models we're currently training; as such, the methodology we employed leaves a lot to be desired. We didn't provide proof ideas or mathematical suggestions to the model during this evaluation; for some solutions, we asked the model to expand upon some proofs, per expert feedback. We also manually facilitated a back-and-forth between this model and ChatGPT for verification, formatting and style. For some problems, we present the best of a few attempts according to human judgement. We are looking forward to more controlled evaluations in the next round! 1stproof.org #1stProof

A flywheel that needs humans to spin it is a treadmill. In 2026, the real turning point is rebuilding RL infra so coding agents can run the loop e2e: implement → debug → run → attribute → report. Shorter cycle time = compounding capability. Full post: hsheng.org/writing/the-re…

@NeurIPSConf “I did it to make my keynote looks better. Nobody at my school told my about DEI and racisim. “ — Excuse given by US profrssor who will be fired from top university Note: Most US faculty who i know are honest and inclusive, except this one.

NeurIPS acknowledges that the cultural generalization made by the keynote speaker today reinforces implicit biases by making generalisations about Chinese scholars. This is not what NeurIPS stands for. NeurIPS is dedicated to being a safe space for all of us. We want to address the comment made during the invited talk this afternoon, as it is something that NeurIPS does not condone and it doesn't align with our code of conduct. We are addressing this issue with the speaker directly. NeurIPS is dedicated to being a diverse and inclusive place where everyone is treated equally.

@moderncpp7 @clu_cheng @NeurIPSConf @drfeifei @jhyuxm @edchi I didn't see the talk, but the images I've seen of the slide seem quite offensive. Such generalizations should have no place in NeurIPS or anywhere else.

Mitigating racial bias from LLMs is a lot easier than removing it from humans! Can’t believe this happened at the best AI conference @NeurIPSConf We have ethical reviews for authors, but missed it for invited speakers? 😡

Sneak peek of our deep learning-based deforestation driver predictions at all (2 million+!) forest loss events in Indonesia from 2012 to 2019! Enabled by the @DescartesLabs platform. Full paper with details coming soon!

My team is growing! Come make a dent in the universe in @Apple Special Projects Group. We're building a unique platform for data-centric deep learning, crowd annotation, and self-supervision. DM me to learn more. #datacentric #AI #ML #MLplatform jobs.apple.com/en-us/details/…

The Google Incentive Mismatch: Problems with Promotion-Oriented Cultures | Warp: warp.dev/blog/problems-…

Andrew Ng is the winner of the 2022 AAAI Robert S. Engelmore Memorial Lecture Award for outstanding contributions to machine learning research and its application to significant real-world problems, and to the democratization of AI through free online education. #AAAI2022

Interested in Using Computer Vision on Street View Imagery to Map Crosswalks of Transit-Oriented Development Station Areas in US? @ced_berkeley PhD Candidate @LiMeiqing can tell you about it w/Stanford Machine Learning Group, @hao_ssr @jeremy_irvin16 @AndrewYNg @NASEMTRB #TRBAM

🧵🧵🧵In our #NeurIPS2021 paper with @hao_ssr and @5harad, we examine how probabilistic predictions, like weather, sports, and election forecasts, evolve over time as new information becomes available, and we introduce a method to predict the predictions. arxiv.org/abs/2106.06515

I am still recovering from yesterday's excitement of defending my thesis. I want to say thank you to everyone who helps me up here and, especially my advisors @5harad and @AndrewYNg, and my spectacular committee Prof. Ram Rajagopal, Prof. @jackiehwang17, and Prof. Rob Jackson.

Happy to share a recent manuscript published in @agformet! Co-led with @sharonz @gavin_mcnicol With students @fredlu_1618 and Vincent Liu Collaborators @etiennefluet Zutao Yang and @sarahelenknox Supervised by Rob Jackson and @AndrewYNg authors.elsevier.com/a/1dNxrcFXJZ1gC 1/5

[6/6] Check out our website for more, including maps of the locations of verified cameras, our computer vision model, and data to use in your own analysis. policylab.stanford.edu/surveillance/

[5/6] We also found more cameras in neighborhoods with higher proportions of non-white residents. This concentration of cameras in majority-minority neighborhoods points to the potential disparate impacts of surveillance technology on communities of color.

[1/6] I'm excited to share our new @AIESConf paper "Surveilling Surveillance: Estimating the Prevalence of Surveillance Cameras with Street View Data" w/ @KenielYao and @5harad arxiv.org/pdf/2105.01764… 🧵

Congrats @OpenAI on really cool image generation result from a text+image prompt! Here's my favorite set of generated examples: