In case you missed itAnthropic says Claude now writes 80% of its codebase, boosting engineer output eightfold as capabilities double every four monthsSD#91|@_SHOLTODOUGLASNL#64|@NATOLAMBERTJC#33|@JACKCLARKSFSB#27|@SLEEPINYOURHATMB#20|@MILES_BRUNDAGE+36 more
#1 VIEWEDSpaceX promotional campaign teases plans to build critical infrastructure across space, connectivity, and AIBY#338|@BILLYUCHENLINEM#74|@ELONMUSK
RISING LIKES3.2kAllen Institute's Finbarr Timbers says DeepSeek V4's Tiananmen Square code censorship points to specialized datasets, but Alexander Doria suggests synthetic rephrasingAD#867|@DORIALEXANDERFI#395|@FINBARRTIMBERS
FASTEST CLIMBINGDominik Kundel, Codex builder, launches /goal mode to run autonomous coding tasks continuously for daysJA#931|@JXNLCORH#804|@ROMAINHUET
pedropark99/zig-book4H AGOProvides the source and examples for a project-based introductory book on Zig.EZ10952.6k stars
ethanhe42/nanoRL6H AGOImplements minimal single-file SFT, DPO, GRPO, and PPO for language model fine-tuning on toy arithmetic tasks.🎭862EH1692108 stars
…tinual-learning-bench6H AGOBenchmarks AI agents on learning from repeated interactions across multi-episode tasks with adaptation metrics.🎭862AZ846127 stars
trycua/cua6H AGOSupplies sandboxes, SDKs, and benchmarks for AI agents to control full desktops on macOS, Linux, and Windows.🎭862DH77317.6k stars
ethanhe42/nanoRL6H AGOImplements minimal single-file SFT, DPO, GRPO, and PPO for language model fine-tuning on toy arithmetic tasks.🎭862EH1692108 stars
…tinual-learning-bench6H AGOBenchmarks AI agents on learning from repeated interactions across multi-episode tasks with adaptation metrics.🎭862AZ846127 stars
trycua/cua6H AGOSupplies sandboxes, SDKs, and benchmarks for AI agents to control full desktops on macOS, Linux, and Windows.🎭862DH77317.6k stars
Yifei-Zuo/Parallax5D AGOImplements parameterized local linear attention mechanisms for efficient language modeling.AC260SY23546 stars