Excited to share that I'll be joining Cornell as an Assistant Professor of CS (@cs_cornell) in January 2023. My research group will focus on optical interconnects and wide area networking. I am looking for students at Cornell! If you are interested in my work, get in touch!

@mdahardy I tried this a while ago when I made https://t.co/xl7bNI4got, looking at the logs, some people have very creative strategies to figure out which model they're talking to! Some of them trigger identifying responses, and some of them figure it out just based on the response time.

How should creative marketplaces be designed when creators train the AI that may displace them? We theoretically analyze this question in a new working paper led by my @CornellMBA PhD student Lijuan Luo in collaboration with @nccyang. Preprint: https://t.co/P2qY3hWMx5

@yifandotqiao Congratulations on the kvcached release! Have you seen our ASPLOS 2025 paper, Aqua, which closely parallels the problems and solutions described in kvcached? 📄 https://t.co/TALDy2dY85 🧵 Thread on how our work aligns:

Our work highlights the potential of AI agents in frontline operations, not as human replacements, but as tools to relieve cognitive burden and improve decisions. Preprint / code at: https://t.co/kTQdMgDANQ cc: @OdedNetzer (@Columbia_Biz), @EvaAscarza (@HarvardHBS), @CornellMBA

Introducing 🛑Stopping Agents: language agents for optimal conversation stopping. We often waste time on conversations that are unlikely to succeed (when selling a product, for example). Stopping Agents save time by stopping such conversations early. https://t.co/u8dzxZbCOo

Introducing CLAMP: : a device, dataset, and model that bring large-scale, in-the-wild multimodal haptics to real robots. Haptics / Tactile data is more than just force or surface texture, and capturing this multimodal haptic information can be useful for robot manipulation.

Why build cool games for humans when you can build them for machines?

Excited to share our preprint: Accelerating AllReduce with a Persistent Straggler 🚀 w/ Eric Ding, @nth_abhishek, Robert Kleinberg, @rachee_singh We design a new algorithm to speed up AllReduce in distributed ML jobs with a persistent straggler GPU 🧵⬇️ https://t.co/VaGaLDCDvY

@nth_abhishek presents our work on server-scale photonic interconnects at @ACMSIGCOMM HotNets! Thanks Sujata for chairing the session!

HotNets is off to an excellent start this morning!

I am hiring a postdoc at Cornell for systems research on next-generation multi-GPU interconnects. If you are about to graduate with a PhD in CS or a related field, email me at rachee@cs.cornell.edu with your CV and a representative publication.

Excited to be back in Pittsburgh @teppercmu this Friday to share ongoing research on "Bounding Human Effort in Augmented Intelligence Systems"! 🌐 https://t.co/BCosydihqg

Four faculty members and four doctoral students from Cornell Bowers CIS are part of the third LinkedIn cohort. The partnership aims to advance research in areas including algorithmic fairness, reinforcement learning, and large language models. https://t.co/KInfLE8zYS

Feeling emo that I am not at this event! Especially since @BjornerNikolaj is DM'ing me that all my favorite people are there!

From the Tortured Traffic Department: Release of #TaylorSwift album leads to music streaming surge, according to @kentikinc traffic stats. https://t.co/GN3opfw8SZ #thetorturedpoetsdepartment

Next up is one of Cornell's brightest, Abhishek Kumar, presenting optical slicing on the WAN using Chisel:

First talk on RAN slicing by Arjun Balasingam (work with Mani and @SuperBahl):

Ying Zhang chairing the session on wide-area and edge session at #NSDI!

https://t.co/HSaeu6R25R

🆕Preprint: Designing Effective Music Excerpts 🌐 https://t.co/o9l6p4LxED Short-form media has created exciting opportunities for music promotion and discovery. But TikTok virality on its own doesn't pay the bills — artists still depend on long-form consumption to make ends meet.