Xin Eric Wang
@xwang_lk
Head of Research @SimularAI. Professor @ UCSB @ucsantabarbara. #Multimodal #Embodied #Agents. AI for Humanity in the long run. ๐ฆ http://ericxw.bsky.social
๐๐๐๐ญ๐ข๐ง๐ ๐๐ฉ๐๐ง๐๐ ๐ข๐ฌ ๐ง๐จ๐ญ ๐๐ฌ ๐ก๐๐ซ๐ ๐๐ฌ ๐ฒ๐จ๐ฎ ๐ญ๐ก๐ข๐ง๐ค. If you don't believe you can compete, you've already lost. Winning starts with mindset. ๐Introducing ๐จ๐๐๐๐ ๐บ2, ๐ญ๐ก๐ ๐ฐ๐จ๐ซ๐ฅ๐'๐ฌ ๐๐๐ฌ๐ญ ๐๐จ๐ฆ๐ฉ๐ฎ๐ญ๐๐ซ-๐ฎ๐ฌ๐ ๐๐ ๐๐ง๐ญ, and the secondโฆ
Introducing Agent S2 โ Our newest open-source AI agent setting new records in computer & smartphone use! We are seeing Agent S2 solve a whole new range of tasks, pushing the boundaries of AI-driven autonomy. ๐ฅ Why itโs special: โ #1 in OSWorld (34.5% accuracy at 50 steps,โฆ
Meta is executing this now.
Unpopular opinion: Big tech should put young minds in AI leadership.
The first man in the Bible? I might know him.
Anyone knows adam?
Lots of US NLP researchers are fleeing #ACL.
Progress for ๐ฎ๐ณ India at ACL, but look at China! Half the papers at ACL from just one country.
When you thought Canada was liberal but it actually has the most conservative visa policy. Heard some incoming faculty couldnโt even make it in and chose other countries instead.
Today I learned a student of mine from China gave up waiting for his Canadian visa after over a year without updates: 1. He was a Vector Scholarship awardee. 2. He had to set aside $20K under the Direct Stream (for faster visa processing), despite being my funded student. 3. Heโฆ
Why don't you just say "this message is for Chinese researchers"? Besides, I am also amazed by your superpower to recognize the ethnicity of anonymous reviewers. Otherwise, how could one just assume a negative review is from a WeChat user?
โI was given an offer that would explode same day. I had to forfeit all of my vested shares earned over my 3.5+ years at Windsurf. I was ultimately given a payout of only 1% of what my shares would have been worth at the time of the deal.โ This is beyond terrible.
Iโve joined Cognition to continue to work on the future of software engineering. I was employee #2 at Windsurf and have worked on AI+code for years. Thereโs never been a more exciting time and place for it than now at Cognition. I had a place at Google DeepMind as part of theโฆ
#NeurIPS reviews and #EMNLP meta reviews are out on the same day. How do you all feel? ๐
So true. Today I had to fight the urge very hard to pull over midway just to write down a research idea that struck me.
True.
@zhenzhangzz (with @xwang_lk) will present his work called "Soft Thinking" -ย a training-free method that emulates human-like "soft" reasoning by generating soft, abstract concept tokens in a continuous concept space. x.com/xwang_lk/statuโฆ
๐๐ถ๐ฎ๐ข๐ฏ๐ด ๐ต๐ฉ๐ช๐ฏ๐ฌ ๐ง๐ญ๐ถ๐ช๐ฅ๐ญ๐บโ๐ฏ๐ข๐ท๐ช๐จ๐ข๐ต๐ช๐ฏ๐จ ๐ข๐ฃ๐ด๐ต๐ณ๐ข๐ค๐ต ๐ค๐ฐ๐ฏ๐ค๐ฆ๐ฑ๐ต๐ด ๐ฆ๐ง๐ง๐ฐ๐ณ๐ต๐ญ๐ฆ๐ด๐ด๐ญ๐บ, ๐ง๐ณ๐ฆ๐ฆ ๐ง๐ณ๐ฐ๐ฎ ๐ณ๐ช๐จ๐ช๐ฅ ๐ญ๐ช๐ฏ๐จ๐ถ๐ช๐ด๐ต๐ช๐ค ๐ฃ๐ฐ๐ถ๐ฏ๐ฅ๐ข๐ณ๐ช๐ฆ๐ด. But current reasoning models remain constrained by discrete tokens, limiting their fullโฆ
๐ง More Thinking, Less Seeing? ๐ We investigate amplified hallucination in multimodal reasoning modelsโwhy current reasoning sometimes hurts perception. Key findings: 1๏ธโฃ Strong reasoning โ strong perception 2๏ธโฃ Longer thoughts โ more hallucination 3๏ธโฃ Moderate reasoning isโฆ
๐ง More Thinking, Less Seeing? ๐ Exploring the Balance Between Reasoning and Hallucination in Multimodal Reasoning Models! Currently many multimodal reasoning models while striving for enhanced reasoning capabilities often neglect the issue of visual hallucinations. Whileโฆ
Silicon Valley drama of the year.
Cognition has signed a definitive agreement to acquire Windsurf. The acquisition includes Windsurfโs IP, product, trademark and brand, and strong business. Above all, it includes Windsurfโs world-class people, whom weโre privileged to welcome to our team. We are also honoringโฆ
An agent-native startup can ship faster than a 100x company. @SimularAI will never let you down.
๐ฌ An agency quoted $20k and 4 weeks for my launch video. I passed. Instead, I spent one weekend, $228 and used an AI creative stack. Result: 2 million views and 3,000 likes. Hereโs the cost breakdown: โ @Google Veo3 - $125 โ @openart_ai - $28 โ @Creati_AI_App - $70 โโฆ
โPapers donโt matter,โ says the one who published hundreds. โPhDs donโt matter,โ says the one whose career was built on theirs. โMoney doesnโt matter,โ says the millionaire / billionaire. Maybe listen to yourself first.
1/ Wait, Bigfoot figured out how to run a startup without drowning in multitasking ๐ It found ๐ฆ๐ถ๐บ๐๐น๐ฎ๐ฟ ๐ฃ๐ฟ๐ผ, the worldโs first production-grade, computer-use agent that runs thousands of steps without a hiccup - working 24/7 so he didnโt have to. So how does Simularโฆ