41. We love scaling inference compute, but it’s costly! Independently sampling parallel attempts might b... We love scaling inference compute, but it’s costly! Independently sampling parallel attempts might be the culprit: it wastes compute rediscovering the same solutions. What if we sc 07/03 01:44 ♡
42. @ @milanganai @katielulula Read the post here! https://t.co/02bfnFx1Rm @milanganai @katielulula Read the post here! ai.stanford.edu/blog/rnb-encor… 💬 0 🔄 2 ❤️ 4 👀 3174 📊 1 ⚡ Powered by xgo.ing 06/30 14:36 ♡
43. N New SAIL Blog post: R&B-EnCoRe: What Should a Robot Actually Think About Before It Acts? @milan... New SAIL Blog post: R&B-EnCoRe: What Should a Robot Actually Think About Before It Acts? @milanganai and @katielulula show how a Vision-Language-Action model can teach itself which 06/30 14:36 ♡
44. B By @kushin_m, @danielwurgaft, @LinasNasvytis, @michaelyli_, @noahdgoodman, and @mcxfrank! By @kushin_m , @danielwurgaft , @LinasNasvytis , @michaelyli_ , @noahdgoodman , and @mcxfrank ! 💬 0 🔄 0 ❤️ 4 👀 1286 📊 1 ⚡ Powered by xgo.ing 06/27 05:04 ♡
45. Auto-psych: a new system where AI agents do the whole science loop themselves -- come up with theori... Auto-psych: a new system where AI agents do the whole science loop themselves -- come up with theories, build experiments, recruit real people online, and improve based on the resu 06/27 04:59 ♡
46. Congrats to @OpenJarvisAI! The future of LM inference runs beyond the datacenter: local by default, ... Congrats to @OpenJarvisAI ! The future of LM inference runs beyond the datacenter: local by default, hybrid by design. Jon Saad-Falcon @JonSaadFalcon Excited to share that @OpenJar 06/26 03:51 ♡
47. Most open agentic datasets target one benchmark. In compute-controlled comparisons, OpenThoughts-Age... Most open agentic datasets target one benchmark. In compute-controlled comparisons, OpenThoughts-Agent-v2 leads at every training set size, and generalizes across seven agentic ben 06/25 15:51 ♡
48. B By the wonderful team @jubayer_hamid, @ifdita_hasan, @michaelyli_, @oshaikh13, @yoonholeee, @DorsaS... By the wonderful team @jubayer_hamid , @ifdita_hasan , @michaelyli_ , @oshaikh13 , @yoonholeee , @DorsaSadigh , @chelseabfinn , and @noahdgoodman 💬 0 🔄 2 ❤️ 5  06/25 06:11 ♡
49. At test time, we wrap LLMs in scaffolds that scale compute every which way -- longer chains, paralle... At test time, we wrap LLMs in scaffolds that scale compute every which way -- longer chains, parallel samples, and aggregation across them. So why do we still train them to use onl 06/25 06:07 ♡
50. B By an incredible team - @jubayer_hamid, @ifdita_hasan, @michaelyli_, @oshaikh13, @yoonholeee, @Dorsa... By an incredible team - @jubayer_hamid , @ifdita_hasan , @michaelyli_ , @oshaikh13 , @yoonholeee , @DorsaSadigh , @chelseabfinn , @noahdgoodman 💬 0 🔄 0 ❤️ 5 ὄ 06/24 07:28 ♡
