October 7, 2025 front

ML Newsnew | past | best | rss | submit

Stories from October 7, 2025 (UTC)

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

1.	Pretraining with hierarchical memories separating long-tail and common knowledge (arxiv.org)
	5 points by dataminer 362 days ago \| hide \| past \| pdf \| discuss
2.	Continuously Augmented Discrete Diffusion Model (arxiv.org)
	4 points by gok 361 days ago \| hide \| past \| pdf \| discuss
3.	HSGM: Hierarchical Segment-Graph Memory for Scalable Long-Text Semantics (arxiv.org)
	3 points by PaulHoule 361 days ago \| hide \| past \| pdf \| discuss
4.	BigBang-Proton, Next-Word-Prediction Is Scientific Multitask Learner (arxiv.org)
	2 points by SSymTech 361 days ago \| hide \| past \| pdf \| 1 comment
5.	Evolution Strategies at Scale: LLM Fine-Tuning Beyond Reinforcement Learning (arxiv.org)
	2 points by ijk 361 days ago \| hide \| past \| pdf \| discuss
6.	The Missing Link Between the Transformer and Models of the Brain (arxiv.org)
	2 points by birriel 362 days ago \| hide \| past \| pdf \| discuss
7.	The Dragon Hatchling (arxiv.org)
	1 point by dboreham 361 days ago \| hide \| past \| pdf \| discuss
8.	Synthetic Bootstrapped Pretraining (arxiv.org)
	1 point by PaulHoule 362 days ago \| hide \| past \| pdf \| discuss