Cédric (@cedcolas) 's Twitter Profile
Cédric

@cedcolas

Building open-ended autotelic agents from socio-cultural interactions

ID: 1014971843315470336

linkhttp://cedriccolas.com calendar_today05-07-2018 20:38:39

432 Tweet

777 Takipçi

462 Takip Edilen

Pierre-Yves Oudeyer (@pyoudeyer) 's Twitter Profile Photo

🧠 One of the key limitation of LLMs today is their lack of metacognition: they were (mostly) not trained to know what they know or don't know, what they can or can't do. 🚀At Inria Flowers team, we're proposing an approach to build metacognition into LLMs: MAGELLAN !

🧠 One of the key limitation of LLMs today is their lack of metacognition:  they were (mostly) not trained to know what they know or don't know, what they can or can't do.

🚀At <a href="/FlowersINRIA/">Inria Flowers team</a>, we're proposing an approach to build metacognition into LLMs: MAGELLAN !
Ġabe Ġrand (@gabe_grand) 's Twitter Profile Photo

New preprint is live! Tweet thread coming 🚧🔜 📅 Excited to present this work in-person: - 4/11: Poster at New England NLP (NENLP) 2025, Yale University (tomorrow!) - 4/27: Oral talk at VerifAI@ICLR 2025, Singapore

Nicolas Yax (@nicolas__yax) 's Twitter Profile Photo

We build a distance matrix from comparing outputs of LLMs to a hundred of different contexts and build maps and trees from this distance matrix. Because PhyloLM only requires sampling very few tokens after a very short contexts the algorithm is particularly cheap to run. 1/10

We build a distance matrix from comparing outputs of LLMs to a hundred of different contexts and build maps and trees from this distance matrix. Because PhyloLM only requires sampling very few tokens after a very short contexts the algorithm is particularly cheap to run. 1/10
Cédric (@cedcolas) 's Twitter Profile Photo

bidding for neurips not a single work on intrinsic motivation or open-endedness, almost no rl, only generic llm papers that are not especially aligned with my work what's going on? is matching broken or is there just nothing interesting for me in the 25k submitted papers?

Peter Henderson (@peterhndrsn) 's Twitter Profile Photo

The next ~1-4 years will be taking the 2017-2020 years of Deep RL and scaling up: exploration, generalization, long-horizon tasks, credit assignment, continual learning, multi-agent interaction! Lots of cool work to be done! 🎮🤖 But we shouldn't forget big lessons from back

The next ~1-4 years will be taking the 2017-2020 years of Deep RL and scaling up: exploration, generalization, long-horizon tasks, credit assignment, continual learning, multi-agent interaction! Lots of cool work to be done! 🎮🤖

But we shouldn't forget big lessons from back
Cédric (@cedcolas) 's Twitter Profile Photo

super cool!! many opportunities to look at the evolution of diverse strategies in multi-player self play, or even maybe the cultural evolution of collaborative strategies?

Cédric (@cedcolas) 's Twitter Profile Photo

new post with a first demo of Tangible Dreams, a primer on CPPNs and a description of my plan to build the hardware! cedriccolas.com/blog/tangible-…

new post with a first demo of Tangible Dreams, a primer on CPPNs and a description of my plan to build the hardware!

cedriccolas.com/blog/tangible-…
Junyi Chu (@junyichu) 's Twitter Profile Photo

Delighted to announce our CogSci '25 workshop at the interface between cognitive science and design 🧠🖌️! We're calling it: Minds in the Making🏺minds-making.github.io Register now! June – July 2024, free & open to the public. (all career stages, all disciplines)