February 20, 2026 front

ML Newsnew | past | best | rss | submit

Stories from February 20, 2026 (UTC)

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

1.	Fast KV Compaction via Attention Matching (arxiv.org)
	73 points by cbracketdash 226 days ago \| hide \| past \| pdf \| 15 comments
2.	Large-scale online deanonymization with LLMs (arxiv.org)
	3 points by DalasNoin 226 days ago \| hide \| past \| pdf \| discuss
3.	Multi-agent cooperation through in-context co-player inference (arxiv.org)
	2 points by gmays 225 days ago \| hide \| past \| pdf \| discuss
4.	Monolith – The research paper behind TikToks algorithm (2022) (arxiv.org)
	2 points by Alifatisk 226 days ago \| hide \| past \| pdf \| discuss
5.	Soft Contamination Means Benchmarks Test Shallow Generalization (arxiv.org)
	2 points by gmays 226 days ago \| hide \| past \| pdf \| discuss
6.	Improving Interactive In-Context Learning from Natural Lang Feedback – DeepMind (arxiv.org)
	2 points by zerop 226 days ago \| hide \| past \| pdf \| discuss
7.	Multi-Turn Intent Detection for LLM and Agent Security (ArXiv) (arxiv.org)
	1 point by sharathr 225 days ago \| hide \| past \| pdf \| 1 comment
8.	Prompt Repetition Improves Non-Reasoning LLMs (arxiv.org)
	1 point by elorant 225 days ago \| hide \| past \| pdf \| discuss
9.	Performance of Deep Material Networks for Multiscale Material Modeling (arxiv.org)
	1 point by PaulHoule 225 days ago \| hide \| past \| pdf \| discuss
10.	The Existence and Behavior of Secondary Attention Sinks (arxiv.org)
	1 point by thw20 226 days ago \| hide \| past \| pdf \| discuss
11.	Prompt Repetition Improves Non-Reasoning LLMs (arxiv.org)
	1 point by beatthatflight 226 days ago \| hide \| past \| pdf \| discuss
12.	When Models Manipulate Manifolds: The Geometry of a Counting Task [pdf] (arxiv.org)
	1 point by vinhnx 226 days ago \| hide \| past \| pdf \| discuss