Top stories in AI today: – Poetiq tops ARC-AGI-2 with Gemini variant
– The Rundown Roundtable: Our AI use cases
– Reverse engineer ads in minutes
– Poetry prompts can bypass AI guardrails
– 4 new AI tools, community workflows, and more Read more: https://
therundown.ai/p/poetiq-crack
s-major-reasoning-benchmark
…
RESEARCH
-

Poetiq Dominates ARC-AGI-2 Benchmark; New AI Tools Released
By
–
-
Sora 2 generations slow, heavy testing on 5.2 suggests imminent release
By
–
Sora 2 generations are taking twice as long as normal today OpenAI is doing heavy testing on 5.2 Release is imminent
-
Engineering Emergent Personalities in LLM Reward Optimization
By
–
There is definitely work going into engineering the "you" simulation – the personality that gets all the rewards in verifiable problems, or all the upvotes from users/judge LLMs, or mimics the responses of SFT, and there is an emergent composite personality from that. My point is
-
BEHAVIOR1K Results Promise Exciting AI Research Journey
By
–
That's precisely why these BEHAVIOR1K results excite me! It's going to be a challenging journey with a clear but worthy North Star for all of us!
-
LLMs as Simulators: Rethinking Prompt Engineering Approaches
By
–
Don't think of LLMs as entities but as simulators. For example, when exploring a topic, don't ask: "What do you think about xyz"? There is no "you". Next time try: "What would be a good group of people to explore xyz? What would they say?" The LLM can channel/simulate many
-
BEHAVIOR Teleoperation System Launches with JoyLo Simulation Data
By
–
We thank @SimovationInc for providing high-quality JoyLo teleoperation data in simulation. BEHAVIOR is built upon @nvidia
’s Omniverse. We thank our sponsors for their generous support! @SimovationInc @nvidia @IMDAsg @StanfordHAI @SchmidtFutures (N/N) -
NeurIPS Workshop: Foundation Models Meet Embodied Agents Challenge
By
–
Don't miss our NeurIPS workshop this Sunday – in about an hour! Come join us and see the winning teams present their solutions. Event: Foundation Models Meet Embodied Agents Challenge
Dec. 7th, 11:00 AM – 1:45 PM PST
Mezzanine Room 15AB, Convention Center We look forward to -

18 Teams Compete in Global AI Challenge Competition
By
–
We received submissions from 18 teams across academia, industry, and independent participants from around the world, including the US, China, the UK, South Korea, and more.
Amazing work from all participants — thank you for making this challenge truly exceptional! (3/N) -

BEHAVIOR: Open-Source Benchmark for Embodied AI and Robotics
By
–
As a reminder, BEHAVIOR is an open-source benchmark, designed to enable and evaluate embodied AI and robotics solutions. It includes everyday household tasks grounded in human needs.
— Fei-Fei Li (@drfeifei) 7 décembre 2025
See the challenge description here: https://t.co/z7YMv9Ih2q (2/N) pic.twitter.com/WyOkYqfPDYAs a reminder, BEHAVIOR is an open-source benchmark, designed to enable and evaluate embodied AI and robotics solutions. It includes everyday household tasks grounded in human needs. See the challenge description here: https://
x.com/drfeifei/statu
s/1962971299246178664
… (2/N) -
DeepMind’s Role in General Intelligence Advances
By
–
Absolute must watch. A reminder of how recent the advances we are seeing now in General Intelligence are (and why there is no reason to believe they shouldn't continue)… and how @demishassabis
, and @GoogleDeepMind are central to everything.