Amine Benhalloum @amine_benh
Agents and Post Training @Meta Superintelligence Labs Joined December 2013-
Tweets148
-
Followers337
-
Following566
-
Likes2K
At #ICLR in Rio 🇧🇷 this week. I’m around all week and happy to meet to chat about agents, research ideas, or just connect. See what we’re working on with Romain’s post below on GAIA2 👇 We’re also hiring in EMEA (and globally across @AIatMeta MSL), so feel free to reach out
I'll be at #ICLR2026 in Rio this week presenting my first PhD paper Gaia2 as an Oral; and the whole team is here too! 🇧🇷 🔹 Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments (arxiv.org/abs/2602.11964) 🎤 Oral: Fri, Apr 24 • 11:06–11:16 AM — Oral Session 3A
In #ICLR2026 this week. Looking forward to great conversations, new ideas, and seeing (old and new) friends. Reach out if you’re around.
I'll be at #ICLR2026 in Rio this week presenting my first PhD paper Gaia2 as an Oral; and the whole team is here too! 🇧🇷 🔹 Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments (arxiv.org/abs/2602.11964) 🎤 Oral: Fri, Apr 24 • 11:06–11:16 AM — Oral Session 3A
I'll be at #ICLR2026 in Rio this week presenting my first PhD paper Gaia2 as an Oral; and the whole team is here too! 🇧🇷 🔹 Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments (arxiv.org/abs/2602.11964) 🎤 Oral: Fri, Apr 24 • 11:06–11:16 AM — Oral Session 3A
Not usually a Meta AI user, but wanted to give them a shot after the latest model release (it's free anyway). So I installed the app on my desktop, and noticed "contemplating" mode (didn't see that on the mobile app btw). When I asked a question, 16 agents simultaneously started working on the question which looks pretty cool!
Go open Meta AI app -> Do something fun -> Create an arcade game -> reply here with the result!
the muse spark API will be coming soon! we have been thrilled with the amount of excitement amongst developers who want to try muse spark inside their agentic harnesses stay tuned!
A strong foundation for what's ahead. We're back :)
1/ today we're releasing muse spark, the first model from MSL. nine months ago we rebuilt our ai stack from scratch. new infrastructure, new architecture, new data pipelines. muse spark is the result of that work, and now it powers meta ai. 🧵
I’m at #NeurIPS2025! Our @Meta MSL Agents team is hiring interns in Paris —DM me if you’re excited to build the next wave of agents, environments, and everything in between.
I'll be @NeurIPSConf in San Diego this week, together with the co-authors of ARE/Gaia2 @mialon_gregoire & @amine_benh . Would love to connect: let’s talk about what’s next for agents!
I am at #NeurIPS2025! I am hiring an intern for our Paris team to succeed @MekalaDheeraj and @ulyanapiterbarg, DM if you want to work on what's next for agents Will also have a look back on Gaia and introduce Gaia2 at the Scaling Environments for Agents workshop on Sunday!
We released ARE and Gaia2 one week ago, time to share some observations and add new models to the leaderboard! huggingface.co/blog/meta-agen…
@xianjun_agi Thank you @xianjun_agi ! It's only the beginning ;)
(🧵) Today, we release Meta Code World Model (CWM), a 32-billion-parameter dense LLM that enables novel research on improving code generation through agentic reasoning and planning with world models. ai.meta.com/research/publi…
🧠Great research from @Meta Superintelligence Labs. Proposes Meta Agents Research Environments (ARE) for scaling up agent environments and evaluations. ARE lets researchers build realistic agent environments, run agents asynchronously, and verify them cleanly. On top of it they release Gaia2, a 1,120 scenario benchmark that stresses search, execution, ambiguity, time pressure, collaboration, and noise, and the results show sharp tradeoffs between raw reasoning and speed or cost. ⚙️ The Core Concepts ARE (Agent Runtime Environment) treats the world as a clocked simulation where everything is an event, the agent runs separately, and interactions flow through tools and notifications. Apps are the tools, environments bundle the apps plus rules, and scenarios package starting state, scheduled events, and a verifier. Traditional old benchmarks froze the world while a model was “thinking.” That made results look clean but ignored the real costs of inference time. In ARE, the world keeps ticking asynchronously. Time passes even while the model is generating, apps can trigger notifications, and other actors may act. So if a model is slow, it directly shows up as missed deadlines in the benchmark. That is exactly why GPT-5 (high) got 79.6 on Search but 0 on Time in default mode. The reasoning quality was excellent, but ARE exposed its inference slowness as a concrete failure mode. When ARE switched to instant mode, stripping out the latency, the model suddenly performed well — proving the bottleneck wasn’t reasoning but raw response time @AIatMeta 🧵 Read on 👇
@ThomasScialom amazing!! 🚀🚀 excited ARE is finally out :)
Did you see that the Agent Research Environment is MCP compatible? -> using any MCP tools with any agent is now completely trivial! Check it out! We've used an LLM agent to 1) move a robot arm remotely 2) depending on real time web search results! :D How to in thread ^^
Wanna upgrade your agent game? With @AIatMeta , we're releasing 2 incredibly cool artefacts: - GAIA 2: assistant evaluation with a twist (new: adaptability, robustness to failure & time sensitivity) - ARE, an agent research environment to empower all! huggingface.co/blog/gaia2
Very cool work from Meta Superintelligence Lab. They are open-sourcing Meta Agents Research Environments (ARE), the platform they use to create and scale agent environments. Great resource to stress-test agents in environments closer to real apps. Read on for more:
Our contribution to the second half of AI 🚀 This has been a joy to build.
🏗️ ARE: scaling up agent environments and evaluations In the LLM+RL era, evals and envs are the bottleneck Happy to release Gaia2, an extensible benchmark for agents aiming to reduce the sim2real gap + ARE, the platform in which Gaia2 is built Enjoy evaluating your agents! 👇
Davis Treybig @TreybigDavis
1K Followers 5K Following Early stage investor at Innovation Endeavors, focused on computing infrastructure, data/AI, and tools for builders.
Derek Cedarbaum @DerekCedarbaum
302 Followers 7K Following Product @ Red 6, #2 FTE | Built the world's first in-air augmented reality system for fighter pilots | 🇺🇸
Matt Wesney @D3VAUX
12K Followers 1K Following human. builder. consciousness. ai systems/safety. music production. 3d artist. building: https://t.co/EAcEDZK5xI
Vaibhavi Singh @__Vaibhavi
784 Followers 363 Following CS grad student @NYU_Courant, on reasoning, retrieval & planning, Prev @Adobe @Salesforce, @EPFL_en scholar
MohammadHossein Rezae... @mhrezaeics
369 Followers 938 Following Post-training Research @ScaleAILabs | Ex Research Intern @StanfordNLP | CS @UArizona
Issa Sugiura @strayer_13
818 Followers 1K Following PhD student @sciencetokyo_en | Intern @SakanaAILabs | Multimodality, Benchmarks
Minh Nhat Nguyen @menhguin
16K Followers 8K Following ai agents @hud_evals (hiring!) | owned @AIHubCentral (1 million users,acq.) ex climate protester🦦I seek Greatness, and to guide humanity through a Golden Age
Ivan Vovk @iyuvovk
251 Followers 376 Following lead ml engineer at @Yandex (agents & reasoning), ex @Huawei, @SamsungResearch, @Skoltech
lovish @louvishh
2K Followers 1K Following founding member @recursive_si | phd @ucl and msl @aiatmeta | previously @googleai. mostly random tweets here.
Young D. Kwon ✈️ ... @YoungDKwon1
479 Followers 2K Following AI Scientist @ Samsung AI | Shipped on-device GenAI to Galaxy flagships (S24 · S25 · S26) | Visiting Scholar & PhD @ Cambridge | ML & Systems Rising Star (2025)
Kyla_Z @Kylabearrrr
8 Followers 127 Following AI Ecosystem & Storytelling | ex-manager of tier-1 tech media | 📍Silicon Valley
Joachim Baumann @joabaum
834 Followers 1K Following Postdoc @StanfordNLP @StanfordAILab / Prev: @MilaNLProc @UZH_en @MPI_IS @CarnegieMellon. CompSocSci, LLMs, algorithmic fairness.
Sanxing Chen @sanxing_chen
498 Followers 518 Following phd-ing @duke_nlp. previously @googledeepmind @msftresearch @uva_ilp. agentic exploration & rag
Devina Jain @letscatanate
52 Followers 195 Following Research @ Lambda | Prev: evals at GM Cruise, UC Berkeley
Houda Nait El Barj @Houda_nait
6K Followers 855 Following AI for Human Flourishing Research @OpenAI https://t.co/uvV7Ifbs0p
Mohamed Hamed @hamed_mo7amed
990 Followers 6K Following I tweet about Computer Vision, Deep Learning, and Artificial Intelligence (AI). Principle AI Engineer. Opinions are my own.
leonson @leonson
562 Followers 2K Following 阅读，探索，思考，写作。注重事实。观点均为个人看法。转推≠赞同。 Reading, exploring, thinking, writing. Stand with the facts. Opinions are my own. RT≠Endorse.
Kate Shapovalenko @kate_shapova
44 Followers 308 Following research data tpm @meta (msl/tbd lab), guest co-instructor (neuro+ai) @mit, ai research @synchroninc, ai/ml @carnegiemellon
Jack Wu @JackTripleU
1K Followers 1K Following Building Products @Meta Superintelligence Labs. Managing AIs and engineers who manage AIs who manage AIs who (I think) write code.
Ender @enderplayerone
137 Followers 871 Following slowly walking down the hall, faster than a canonball
Clayton Thorrez @cthorrez
1K Followers 3K Following Rating systems and paired comparison experimentation enjoyer @arena
Minh Nguyen @MinhNguyen1494
382 Followers 2K Following
Mircea Mironenco @mirceamironenco
105 Followers 3K Following
Mu Cai @MuCai7
3K Followers 1K Following Research @thinkymachines | Previous: multimodal, agents @GoogleDeepMind
Scott McCrae @scottymccrae
233 Followers 1K Following superintelligence @meta. helping machines learn :)
Tristan Zajonc @tristanzajonc
2K Followers 2K Following Cofounder/CEO at @Continual_AI. AI, data, startups, economics. Formerly @Cloudera, @SensePlatform, @Harvard.
Bruno De Martino @bdmartino
450 Followers 803 Following AI agents @nubank. Prev: @instagram, startups. Computer Science @stanford. 🇧🇷🇺🇸🇮🇹
Andrew Qian @AndrewQfeeder
9 Followers 613 Following
Pierre Chambon @PierreChambon6
838 Followers 2K Following NLP/Code Generation PhD at FAIR (Meta AI) and INRIA - previously researcher at Stanford University - MS Stanford 22’ - Centrale Paris P2020
Jun @junasi8
40 Followers 5K Following
nvd @imnojan
145 Followers 988 Following
Stephane Kasriel @skasriel
23K Followers 3K Following VP at Meta FAIR, Meta Fundamental AI Research. Follow us at @aiatmeta.
Gabriel Synnaeve @syhw
17K Followers 1K Following Nerd & Dad. RL & CodeGen research since before it was cool.
Jaydeep Mankikar @MankikarJaydeep
1 Followers 1K Following
Kate Shapovalenko @kate_shapova
44 Followers 308 Following research data tpm @meta (msl/tbd lab), guest co-instructor (neuro+ai) @mit, ai research @synchroninc, ai/ml @carnegiemellon
Joachim Baumann @joabaum
834 Followers 1K Following Postdoc @StanfordNLP @StanfordAILab / Prev: @MilaNLProc @UZH_en @MPI_IS @CarnegieMellon. CompSocSci, LLMs, algorithmic fairness.
Ivan Vovk @iyuvovk
251 Followers 376 Following lead ml engineer at @Yandex (agents & reasoning), ex @Huawei, @SamsungResearch, @Skoltech
MohammadHossein Rezae... @mhrezaeics
369 Followers 938 Following Post-training Research @ScaleAILabs | Ex Research Intern @StanfordNLP | CS @UArizona
Davis Treybig @TreybigDavis
1K Followers 5K Following Early stage investor at Innovation Endeavors, focused on computing infrastructure, data/AI, and tools for builders.
Hardik Bhatnagar @hrdkbhatnagar
551 Followers 260 Following Building PostTrainBench, current @MATSprogram Evals, Long horizon, Interp, Safety | PhD @ Max Planck, Tübingen Prev: @MSFTResearch
lovish @louvishh
2K Followers 1K Following founding member @recursive_si | phd @ucl and msl @aiatmeta | previously @googleai. mostly random tweets here.
Devina Jain @letscatanate
52 Followers 195 Following Research @ Lambda | Prev: evals at GM Cruise, UC Berkeley
Houda Nait El Barj @Houda_nait
6K Followers 855 Following AI for Human Flourishing Research @OpenAI https://t.co/uvV7Ifbs0p
Mark Chen @markchen90
74K Followers 353 Following Chief Research Officer at @OpenAI. Coach for the USA IOI Team.
Pierre Chambon @PierreChambon6
838 Followers 2K Following NLP/Code Generation PhD at FAIR (Meta AI) and INRIA - previously researcher at Stanford University - MS Stanford 22’ - Centrale Paris P2020
Tristan Zajonc @tristanzajonc
2K Followers 2K Following Cofounder/CEO at @Continual_AI. AI, data, startups, economics. Formerly @Cloudera, @SensePlatform, @Harvard.
Scott McCrae @scottymccrae
233 Followers 1K Following superintelligence @meta. helping machines learn :)
Mu Cai @MuCai7
3K Followers 1K Following Research @thinkymachines | Previous: multimodal, agents @GoogleDeepMind
Jack Wu @JackTripleU
1K Followers 1K Following Building Products @Meta Superintelligence Labs. Managing AIs and engineers who manage AIs who manage AIs who (I think) write code.
Shishir Patil @shishirpatil_
4K Followers 1K Following CS PhD @ UC Berkeley. Creator of Gorilla, GoEx, RAFT, OpenFunctions and Berkeley Function Calling Leaderboard. Previously researcher @GoogleAI @MSFTResearch
Prakhar Agarwal @prakhar5050
2K Followers 5K Following @Meta Superintelligence Labs | Prev. Research @openai, @apple. Grad Student at Univ. of Washington (@uwcse) , RF at @MSFTResearch | YR @HLForum
Megha Dasgupta Purkay... @meghadasgupta
481 Followers 529 Following Member of Recruiting Staff @ Microsoft AI
William Sparks @WilliamWSparks
672 Followers 4K Following Google DeepMind | Recruiting and Analytics @Google F/@Yieldmo, F/ @ZocDoc, @Babson Beaver, @NYUStern #MBA My tweets are my own.
Vishrav Chaudhary @vishrav
778 Followers 747 Following Researcher at Superintelligence Lab @MetaAI. Ex- @Microsft Turing @MetaAI @LTIatCMU alum.
Ashwin Vaswani @ashwin_vaswani
2K Followers 6K Following Research Scientist @GoogleDeepMind | Prev: @CarnegieMellon | @GoogleIndia | APPCAIR, @BITSPilaniGoa | @qtimlab, Harvard
Nikos Pantelaios @PantelaiosNikos
84 Followers 246 Following AI Research Scientist @ Meta, FAIR, Meta Superintelligence Labs Post-Training, RL, Muse Spark
Yixin Lin @yixin_lin_
2K Followers 7K Following something new. prev: embodied AI @GoogleDeepMind, FAIR/@AIatMeta, Google Brain.
Stephane Kasriel @skasriel
23K Followers 3K Following VP at Meta FAIR, Meta Fundamental AI Research. Follow us at @aiatmeta.
Mathieu @Mathieu_Rita
231 Followers 277 Following Research Scientist @AIatMeta ex: INRIA-MSR | @CoML_ENS | @Polytechnique Llama3 - RL fine-tuning - Emergent communication
Alaa El-Nouby @alaa_nouby
827 Followers 439 Following Research Scientist at Meta Superintelligence Labs. Previous: @Apple, @Inria, @MSFTResearch, @VectorInst and @UofG
Zhiqing Sun @EdwardSun0909
20K Followers 1K Following Lead agent research @Meta MSL TBD Lab. previously posttraining/agent research @OpenAI. CS PhD @LTIatCMU
Qian Huang @qhwang3
14K Followers 331 Following prev @xai | CS PhD student @StanfordAILab (on leave)
Deepak Nathani @deepaknathani11
943 Followers 1K Following PhD Student @UCSBNLP | Prev: @AIatMeta | @AWS AI | @GoogleAI India | @IITHyderabad
Claude @claudeai
1.5M Followers 2 Following Claude is an AI assistant built by @anthropicai to be safe, accurate, and secure. Talk to Claude on https://t.co/ZhTwG8d1e5 or download the app.
Sungmin Cha @_sungmin_cha
1K Followers 326 Following Research Scientist at Meta | Formally Faculty Fellow @nyuniversity | PhD @SeoulNatlUni
Zach Xu @nehzux
151 Followers 1K Following Working on Language Model @Meta Superintelligence Labs | PhD @UChicago
Gradium @GradiumAI
4K Followers 1 Following The voice layer for modern apps and agents. Real-time, scalable voice APIs: TTS, STT, turn-taking & voice cloning. Devs: build → https://t.co/r5CdNClhI5
Jason Weston @jaseweston
15K Followers 898 Following Senior Director & RS @Meta + Visiting Prof NYU | OG in LLMs | Pretrain+Finetune in 2008+ | 151k+ citations | Current: Self-Improving & Co-Improving AI
Dhruv Batra @DhruvBatra_
21K Followers 732 Following Co-founder & Chief Scientist @yutori_ai. Prev: Senior Director leading FAIR Embodied AI @MetaAI and Professor @GeorgiaTech.
Tri Dao @tri_dao
42K Followers 657 Following Asst. Prof @PrincetonCS, Chief Scientist @togethercompute. Machine learning & systems.
Susan Zhang @suchenzang
47K Followers 1K Following @ Google Deepmind. Past: @MetaAI, @OpenAI, @unitygames, @losalamosnatlab, @Princeton etc. Always hungry for intelligence. Only my opinions stored here.
Grad @Grad62304977
9K Followers 3K Following
Teknium 🪽 @Teknium
106K Followers 6K Following Cofounder and Lead Engineer - Hermes Agent @NousResearch, prev @StabilityAI Github: https://t.co/LZwHTUFwPq HuggingFace: https://t.co/sN2FFU8PVE
Lautaro Gesuelli @lgesuelli_p
998 Followers 2K Following founding engineer @taste_ai_ | building llmnative 🇦🇷 | professorYou might like
