hlfshell · @hlfshell
74 followers · 324 posts · Server hachyderm.ioDue to travel I've had time to sit down and start working on my large research paper to-read list. The 2 I chose to start with were incredible:
First up is SayCan- using RL to guide LLMs to controlling robots in a complex kitchen environment
https://say-can.github.io/
...then we haveGenerative Agents: Interactive Simulacra of Human Behavior- where authors used LLMs to control multiple agents and had them simulate a small town of unique personalities
https://arxiv.org/abs/2304.03442
Rene Schulte · @rschu
594 followers · 287 posts · Server arvr.socialWOAH! 🤯 The first autonomous vision-based drone that beats human world champions in head-to-head races.
They use Reinforcement Learning to achieve this groundbreaking mobile #robotics milestone. 🤖
Open access paper in Nature: https://www.nature.com/articles/s41586-023-06419-4
Author post: https://twitter.com/davsca1/status/1696938013421429111
Full video: https://www.youtube.com/watch?v=fBiataDpGIo
Cool stuff but also terrifying implications if you see how these FPS drones are right now used in active conflicts.
#robotics #ai #rl #cv #deeplearning
New Submissions to TMLR · @tmlrsub
205 followers · 743 posts · Server sigmoid.social
RLTF: Reinforcement Learning from Unit Test Feedback
Andreas Preissig · @bahntiger
10 followers · 30 posts · Server zug.networkAktuell fährt die #RheinNeckar S6 zwischen #RL und #RM über das Gütergleis der Strecke 3401. Wenn da noch jemand Strecken und Weichen sammeln will...
@nordkommission zum Beispiel? Bevor Du heute nur traurig am Flughafen sitzt...
hlfshell · @hlfshell
67 followers · 247 posts · Server hachyderm.ioI finally finished my writeup on utilizing PPO to control a robotic arm to attempt to solve a pick and place problem.
https://hlfshell.ai/posts/ppo-pick-and-place/
In the post I discuss my successes, failures, how everything works, and how I debugged the problem.
It's my first attempt at an in depth tech blogpost.
#robotics #rl #reinforcementlearning #ai
Thibaut Bulle Darchi 🫧 · @Thibulle
10 followers · 97 posts · Server ludosphere.fr
Paré à se faire rouler dessus par les autres streamers pour l’Opensub de RTBF_Ixpé avec l’incroyable participation de Prof_Poncho 👌
#stream #twitch #jeuxvideo #rl #belgium #Belgique
New Submissions to TMLR · @tmlrsub
198 followers · 689 posts · Server sigmoid.social
Bridging the Gap Between Offline and Online Reinforcement Learning Evaluation Methodologies
Mihai Lazarescu · @mtl
1 followers · 139 posts · Server techhub.social
#ESC is a #ZeroShot agent that uses #ChatGPT and #SoftLogic to navigate in 3D environments with #CommonSense and #WorldKnowledge. #AI #NLP #RL
https://www.marktechpost.com/2023/07/22/researchers-from-uc-santa-cruz-and-samsung-introduce-esc-a-zero-shot-object-navigation-agent-that-leverages-commonsense-in-llms-like-chatgpt-for-navigation-decisions/
#esc #zeroshot #chatgpt #softlogic #commonsense #worldknowledge #AI #nlp #rl
Games Japan · @games
5 followers · 143 posts · Server wakoka.com
https://www.wacoca.com/games/683402/ 敵避けて右に曲げるフリップリセット最強最強最強最強最強最強最強最強 【ロケットリーグ】 #shorts #Bloom #Epic #freestyle #PS4 #PS4GAMES #rl #RocketLeague #SSL #steam #Switch #TeamBloom #TeamNytro #Xbox #YMTO #YmtoS #おもしろ #グランド #サッカー #スーパープレイ #フリースタイル #フリスタ #プロ #ヤマキンTV #ヤマト #ロケットリーグ #ロケリ #大和 #山芋たいこく
#shorts #bloom #epic #freestyle #ps4 #ps4games #rl #rocketleague #ssl #steam #switch #teambloom #teamnytro #xbox #ymto #ymtos #おもしろ #グランド #サッカー #スーパープレイ #フリースタイル #フリスタ #プロ #ヤマキンtv #ヤマト #ロケットリーグ #ロケリ #大和 #山芋たいこく
Games Japan · @games
1 followers · 6 posts · Server wakoka.com
https://www.wacoca.com/games/677305/ 新モードで面白い動画撮ろうとしたら日常動画になっちゃったよ。【ロケットリーグ】 #Bloom #Epic #freestyle #PS4 #PS4GAMES #rl #RocketLeague #SSL #steam #Switch #TeamBloom #TeamNytro #Xbox #YMTO #YmtoS #おもしろ #グランド #サッカー #スーパープレイ #フリースタイル #フリスタ #プロ #ヤマキンTV #ヤマト #ロケットリーグ #ロケリ #大和 #山芋たいこく
#bloom #epic #freestyle #ps4 #ps4games #rl #rocketleague #ssl #steam #switch #teambloom #teamnytro #xbox #ymto #ymtos #おもしろ #グランド #サッカー #スーパープレイ #フリースタイル #フリスタ #プロ #ヤマキンtv #ヤマト #ロケットリーグ #ロケリ #大和 #山芋たいこく
Todor Stoyanov · @todor
49 followers · 26 posts · Server sigmoid.socialLooking for a #PhD position in #ML and #robotics? I have an open position focusing on learning full-body manipulation affordances. We will look into #NeRFs and try both supervised and #RL approaches for learning on a real mobile YuMi robot. To apply please go through our online system here:
https://www.oru.se/english/career/available-positions/job/?jid=20230237
Thanks for the re-tooth 😊
Andreas Preissig · @bahntiger
1 followers · 2 posts · Server zug.network
Heute mal ein wenig Umleitungen mitfahren. Als erstes mal einer der letzten Fernverkehrszüge aus #RL, heute passend bis Esslingen...
Antjh1981 · @Antjh
0 followers · 87 posts · Server mas.to#McLelland Ends #RL Career Early!
#Bulls Sign 2!
#Wire Womens Player Stood Down!
#TimeIsNow #GROWTheGame #GrowRugbyLeague #RugbyLeague #RLFamily #SuperLeague #NRL #NARL #RFL #IRL #FFR #USARL #PDRL #RLIF #RLWC2021 #RespectTheREF
#respecttheref #RLWC2021 #rlif #pdrl #usarl #ffr #irl #RFL #narl #nrl #SuperLeague #rlfamily #rugbyleague #growrugbyleague #growthegame #timeisnow #wire #bulls #rl #mclelland
Aran Komatsuzaki · @aran
920 followers · 871 posts · Server sigmoid.socialJumanji: a Diverse Suite of Scalable Reinforcement Learning Environments in JAX
https://arxiv.org/abs/2306.09884
RT @instadeepai@twitter.com
1/ Exciting news! Our team has just released a major update of Jumanji, our suite of diverse and challenging #RL environments written in #JAX 🔥 Check it out now and take your research to the next level 🚀
⭐ Github: http://tinyurl.com/code-jumanji
📚 Doc: http://tinyurl.com/doc-jumanji
🐦🔗: https://twitter.com/instadeepai/status/1638571324992888832
Brandon Rohrer · @brohrer
1230 followers · 1055 posts · Server recsys.socialDoes one #ReinforcementLearning algorithm perform better than another? It is notoriously hard to answer this question well, even when your intentions are good. The authors here give recommendations that are clear, sensible, and practicable.
To me, this paper is the pinnacle of scholarship. Statistics-savvy #RL researchers extend a hand to those that are less so, without talking down or getting aggressively technical. This is the work of bridge builders.
Thibaut Bulle Darchi 🫧 · @Thibulle
6 followers · 75 posts · Server ludosphere.fr
Ce soir, on regardera le #PlayStationShowcase et samedi, c'est le GRAND #ThibulleChallenge 👩🎤
👉ttv/thibulle
#Stream #Twitch #jeuxvideo #RL #RocketLeague #Sony #Playstation #twitchbe #twitchfr #Lake #Indiegame
#playstationshowcase #thibullechallenge #stream #twitch #jeuxvideo #rl #rocketleague #sony #Playstation #twitchbe #twitchfr #lake #indiegame
652349 · @652349
126 followers · 4020 posts · Server ruhr.social
Ben Waber · @bwaber
534 followers · 1588 posts · Server hci.socialNext was a great talk by Anne Collins on bridging #cognition, #neuroscience, and computation in #RL at the Learning Salon. After some bombastic claims that "RL is all you need" to explain cognition, Collins and the broader group dissect what's missing from this picture https://www.youtube.com/watch?v=YLbZh-bH8V0 (3/9) #ReinforcementLearning
#cognition #neuroscience #rl #reinforcementlearning
5h15h · @shish
95 followers · 616 posts · Server techhub.social#OpenAI uses #ReinforcementLearning from human feedback (#RLHF), an established technique, to enhance the safety, usefulness, and alignment of its models https://openai.com/research/instruction-following
#openai #reinforcementlearning #rlhf #AI #genai #generativeAI #chatgpt #gpt #rl
hlfshell · @hlfshell
40 followers · 132 posts · Server hachyderm.ioWork in progress reinforcement learning project. One block, no "blocker bar" blocking the goal. Each colored zone awards points for a shape being pushed into it, but a specific shape gets extra points in particular zones.
All trained w/ PPO, about 12 million timesteps.
#robotics #reinforcementlearning #deeplearning #rl