March 13, 2026 front

ML Newsnew | past | best | rss | submit

Stories from March 13, 2026 (UTC)

Go back a day, month, or year. Go forward a day.

1.	AutoHarness: Improving LLM agents by automatically synthesizing a code harness (arxiv.org)
	10 points by simonpure 205 days ago \| hide \| past \| pdf \| discuss
2.	End-to-End Hardware-Driven Graph Preprocessing for Enhanced GNN Performance (arxiv.org)
	5 points by PaulHoule 205 days ago \| hide \| past \| pdf \| discuss
3.	Lost in Backpropagation: The LM Head Is a Gradient Bottleneck (arxiv.org)
	4 points by famouswaffles 205 days ago \| hide \| past \| pdf \| discuss
4.	Private LLM Inference on Consumer Blackwell GPUs (arxiv.org)
	3 points by rohansood15 205 days ago \| hide \| past \| pdf \| discuss
5.	LDP: Identity-Aware Routing for Multi-Agent LLMs – 37% Less Tokens (arxiv.org)
	2 points by prakashsunil 205 days ago \| hide \| past \| pdf \| discuss
6.	We Automated RL Environment Engineering for $10 (arxiv.org)
	2 points by milkkarten 206 days ago \| hide \| past \| pdf \| discuss
7.	When Models Examine Themselves: Vocabulary-Activation Correspondence (arxiv.org)
	1 point by tcbrah 205 days ago \| hide \| past \| pdf \| discuss
8.	Native CLI scaffolds consistently outper-form OpenCode when using the same model (arxiv.org)
	1 point by xdotli 205 days ago \| hide \| past \| pdf \| 1 comment