Meena Jagadeesan

I'm so excited to be joining @Penn as an Assistant Professor in CS (@CIS_Penn) in Fall 2026! I’ll be working on machine learning ecosystems, aiming to steer how multi-agent interactions shape performance trends and societal outcomes. I’ll be recruiting PhD students this cycle!

New paper & surprising result. LLMs transmit traits to other models via hidden signals in data. Datasets consisting only of 3-digit numbers can transmit a love for owls, or evil tendencies. 🧵

Today we're releasing Community Alignment - the largest open-source dataset of human preferences for LLMs, containing ~200k comparisons from >3000 annotators in 5 countries / languages! There was a lot of research that went into this... 🧵

My PhD materials are now available! Dissertation: arxiv.org/abs/2506.23123 Slides: drive.google.com/file/d/13N2FRW… Folks should read the acknowledgements since so many people have been so important to me along this journey!

Our special feature in PNAS released earlier this week: Collective Artificial Intelligence and Evolutionary Dynamics - you can find the collection of papers here: pnas.org/topic/569 Enjoy! The Collective Artificial Intelligence and Evolutionary Dynamics Special Feature…

New alignment theory paper! We present a new scalable oversight protocol (prover-estimator debate) and a proof that honesty is incentivised at equilibrium (with large assumptions, see 🧵), even when the AIs involved have similar available compute.

New on the Anthropic Engineering blog: how we built Claude’s research capabilities using multiple agents working in parallel. We share what worked, what didn't, and the engineering challenges along the way. anthropic.com/engineering/bu…

My @a16zcrypto colleagues and I just shared a roundup on the ways crypto will organize, complement – and occasionally tame – modern AI. My takes: 🧵👇 ➡️ You shouldn't have to remind an LLM like Claude of your preferred programming languages each time you open a new shell – and…

Can we systematically generalize AR "word models" into "world models”? Our CVPR 2025 paper introduces a unified, general framework designed to model real-world, multi-agent interactions by disentangling task-specific modeling from behavior prediction.

Got a shiny parting gift from @ucsd_cse, the 2025 Doctoral Award for Excellence in Research! 🎓✨ Grateful for the recognition of my work on Incentive-Aware Trustworthy AI :) Awards certainly weren’t the goal… but hey, I’ll take the trophy and run! 😄

We’re organizing a tutorial on the Economics of Data as part of @AcmSIGecom EC ’25! 🗓️ June 18 | 🕑 2–4 PM ET | 📍Zoom Registration is free but mandatory! Details and registration link are available on the tutorial webpage!👇

Listen to a conversation between the team at @cursor_ai and Anthropic:

Our paper Statistical Collusion by Collectives on Learning Platforms has been selected for an oral presentation at #ICML2025! 🎤 Thanks to my supervisors 🙏 Looking forward to presenting! 💥

I’m very excited to be joining MIT Biological Engineering @MITdeptofBE as an Assistant Professor and the Ragon Institute @ragoninstitute as a member in January 2026! I will be recruiting students and postdocs (see more info below).

is the rational agent with perfect information a suitable model for studying strategic classification? we show degeneracies of this standard model and search for alternative microfoundations.👥 Hope to see you tonight at the #ICML2021 poster session! w/@mjagadeesan25 and @mrtz