February 6, 2025 front

ML Newsnew | past | best | rss | submit

Stories from February 6, 2025 (UTC)

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

1.	Pre-Trained Large Language Models Use Fourier Features for Addition (2024) (arxiv.org)
	149 points by Kye on Feb 6, 2025 \| hide \| past \| pdf \| 40 comments
2.	High-Fidelity Simultaneous Speech-to-Speech Translation (arxiv.org)
	6 points by exgrv on Feb 6, 2025 \| hide \| past \| pdf \| 1 comment
3.	Understanding Why Adam Outperforms SGD: Gradient Heterogeneity in Transformers (arxiv.org)
	3 points by fofoz on Feb 6, 2025 \| hide \| past \| pdf \| discuss
4.	Develop AI Agents for System Engineering in Factorio (arxiv.org)
	3 points by Jimmc414 on Feb 6, 2025 \| hide \| past \| pdf \| discuss
5.	The Hyperfitting Phenomenon: Sharpening and Stabilizing LLMs (arxiv.org)
	3 points by superidiot1932 on Feb 6, 2025 \| hide \| past \| pdf \| discuss
6.	Demystifying Long Chain-of-Thought Reasoning in LLMs (arxiv.org)
	2 points by sebg on Feb 6, 2025 \| hide \| past \| pdf \| discuss
7.	Optimizing LLM Persuasion with Personalization and Fabricated Statistics (arxiv.org)
	2 points by PaulHoule on Feb 6, 2025 \| hide \| past \| pdf \| discuss
8.	LIMO: Less Is More for Reasoning (arxiv.org)
	2 points by maksimur on Feb 6, 2025 \| hide \| past \| pdf \| discuss
9.	SmolLM2: When Smol Goes Big – Data-Centric Training of a Small Language Model (arxiv.org)
	1 point by rahimnathwani on Feb 6, 2025 \| hide \| past \| pdf \| discuss