Evan Wang

@xAI prev @Caltech @scale_AI @weHRTyou @umdcs

[Evan Wang @evanzwangg]

mm rocket bro 🚀

[SpaceXAI @SpaceXAI]

Announcing Grok 4.5, our first model trained specifically for coding and agents. It was trained with Cursor and offers frontier intelligence at leading speeds and cost efficiency.

https://t.co/i8HpU7w64k

[Evan Wang @evanzwangg]

@ArashBidgoli @VastoLorde95 heyyy

[Evan Wang @evanzwangg]

@HonghuaZhang2 😭 😭 it was great working with you! gbab

[Evan Wang @evanzwangg]

@jiayi_pirate @billyuchenlin @skcd42 @veggie_eric truly an irreplaceable legend 🙏🚀

[Evan Wang @evanzwangg]

@jiayi_pirate it was a great pleasure working with you in all the warrooms and seeing you around the office Jiayi! we'll miss you dearly

[Evan Wang @evanzwangg]

@r_ahulravi @xai @Guodzh @archanfel_anoth i’ll miss u man it’s been great having you around 🥲

[Evan Wang @evanzwangg]

@Yuhu_ai_ was an honor to work with you Tony! thx for all the warrooms! 🚀

[Evan Wang @evanzwangg]

@YifeiZhou02 @xai we will miss you yifei!! 🚀

[Evan Wang @evanzwangg]

you love to see it 🐦‍⬛

[SpaceXAI @SpaceXAI]

Introducing Grok 4 Fast, a multimodal reasoning model with a 2M context window that sets a new standard for cost-efficient intelligence.

Available for free on https://t.co/AnXpIEOhOD, https://t.co/53pltypvkw, iOS and Android apps, and OpenRouter.

https://t.co/3YZ1yVwueV

[Evan Wang @evanzwangg]

🐆 💨 we do a bit of coding

[SpaceXAI @SpaceXAI]

Introducing Grok Code Fast 1, a speedy and economical reasoning model that excels at agentic coding.

Now available for free on GitHub Copilot, Cursor, Cline, Kilo Code, Roo Code, opencode, and Windsurf.

https://t.co/3tMbmLbxOP

[Evan Wang @evanzwangg]

✌️

[Daniel @nearlydaniel]

War Room squad locked in

[Evan Wang @evanzwangg]

good stuff grok 🚀
https://t.co/4mfdh8X01S

[Rez Havaei @HavaeiRez]

LLMs are being deployed in high-stakes environments—and the potential impact of failure is colossal. A jailbroken AI could leak your customer data, financial records, or enable catastrophically harmful actions. At @gen_analysis we have compiled the definitive guide to understand and defend against these threats-- The LLM Jailbreak Cookbook.

Here is an Overview:

(1/8)

[Evan Wang @evanzwangg]

Delighted to announce that PlanSearch has been accepted to ICLR 2025!! 😁😁

see you in singapore 🫡

[Evan Wang @evanzwangg]

A 20% boost on a metric is rare, especially when it’s code generation 🥱

PlanSearch, our new search method based on diverse plans, outperforms baselines by huge margins. It's not just a search method, but also a philosophy

How are these numbers achieved? can they be predicted? 🤨🤨

[Evan Wang @evanzwangg]

thanks for having me!! was great seeing everyone again 😁

[Furong Huang @furongh]

Our very own @evanzwangg visited us back at UMD today and gave an awesome talk. Check out his paper here to see how planning improves pass@k significantly for coding problems: https://t.co/rE505uLqsS

[Evan Wang @evanzwangg]

@ChowdhuryNeil @hughbzhang @ellev3n11 @squeakymouse777 @vaskar_n @SeanHendryx @summeryue0 thanks neil! my friends and i agree that this is a classic neil response 🙂‍↕️🙂‍↕️

[Evan Wang @evanzwangg]

A 20% boost on a metric is rare, especially when it’s code generation 🥱

PlanSearch, our new search method based on diverse plans, outperforms baselines by huge margins. It's not just a search method, but also a philosophy

How are these numbers achieved? can they be predicted? 🤨🤨

[Evan Wang @evanzwangg]

@lunw1024 @hughbzhang @ellev3n11 @squeakymouse777 @vaskar_n @SeanHendryx @summeryue0 thanks alan 🫡🫡

[Evan Wang @evanzwangg]

@lupdestepa @hughbzhang @summeryue0 @squeakymouse777 @ellev3n11 @SeanHendryx described in methods and appendices sections 🙂‍↕️🙂‍↕️