February 11, 2025 front

ML Newsnew | past | best | rss | submit

Stories from February 11, 2025 (UTC)

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

1.	LLMs can teach themselves to better predict the future (arxiv.org)
	176 points by bturtel on Feb 11, 2025 \| hide \| past \| pdf \| 86 comments
2.	Time to act on the risk of efficient personalized text generation (arxiv.org)
	57 points by Jimmc414 on Feb 11, 2025 \| hide \| past \| pdf \| 34 comments
3.	Frontier AI systems have surpassed the self-replicating red line (arxiv.org)
	3 points by rahton on Feb 11, 2025 \| hide \| past \| pdf \| 4 comments
4.	The Curse of Depth in Large Language Models [pdf] (arxiv.org)
	2 points by mkaic on Feb 11, 2025 \| hide \| past \| pdf \| discuss
5.	Training LLMs to Reason Efficiently (arxiv.org)
	2 points by omarsar on Feb 11, 2025 \| hide \| past \| pdf \| discuss
6.	Matryoshka Quantization (arxiv.org)
	2 points by fzliu on Feb 11, 2025 \| hide \| past \| pdf \| discuss
7.	Turning Up the Heat: Min-P Sampling for Creative and Coherent LLM Outputs (arxiv.org)
	1 point by Der_Einzige on Feb 11, 2025 \| hide \| past \| pdf \| discuss
8.	Resurrecting saturated LLM benchmarks with adversarial encoding (arxiv.org)
	1 point by bearseascape on Feb 11, 2025 \| hide \| past \| pdf \| discuss
9.	Deep Networks Always Grok and Here Is Why (arxiv.org)
	1 point by ziofill on Feb 11, 2025 \| hide \| past \| pdf \| discuss