New Links | ML News

ML Newsnew | past | best | rss | submit

1621.	Pulse: Decentralized RL training centralized speed (100x weight sync reduction) (arxiv.org)
	2 points by synapz_org 243 days ago \| hide \| past \| pdf \| discuss
1622.	Who's in Charge? Disempowerment Patterns in Real-World LLM Usage (arxiv.org)
	3 points by 7777777phil 243 days ago \| hide \| past \| pdf \| discuss
1623.	Are Pre-Trained Convolutions Better Than Pre-Trained Transformers? (2021) (arxiv.org)
	2 points by fzliu 243 days ago \| hide \| past \| pdf \| discuss
1624.	Pendulum: A Benchmark for Assessing Sycophancy in MLLM's (arxiv.org)
	1 point by onestay42 243 days ago \| hide \| past \| pdf \| discuss
1625.	Exploration Posteriors for Generative Modeling Using Only Negative Rewards (arxiv.org)
	1 point by numeri 243 days ago \| hide \| past \| pdf \| discuss
1626.	Revisiting Disaggregated LLM Serving for Performance and Energy Implications (arxiv.org)
	1 point by PaulHoule 244 days ago \| hide \| past \| pdf \| discuss
1627.	Linear representations in LLMs can change dramatically over a conversation (arxiv.org)
	5 points by gmays 244 days ago \| hide \| past \| pdf \| discuss
1628.	Language-Related Ideological Divergence in LLM Analysis of Political Documents (arxiv.org)
	1 point by PaulHoule 244 days ago \| hide \| past \| pdf \| discuss
1629.	Reasoning Models Generate Societies of Thought (arxiv.org)
	2 points by PaulHoule 244 days ago \| hide \| past \| pdf \| discuss
1630.	Strategies of cooperation and defection in five large language models (arxiv.org)
	1 point by PaulHoule 244 days ago \| hide \| past \| pdf \| discuss
1631.	A Pragmatic VLA Foundation Model (arxiv.org)
	1 point by mountainview 244 days ago \| hide \| past \| pdf \| discuss
1632.	PaperBanana: Automating Academic Illustration for AI Scientists (arxiv.org)
	1 point by fzliu 244 days ago \| hide \| past \| pdf \| discuss
1633.	Power Aware Dynamic Reallocation for Inference (arxiv.org)
	3 points by PaulHoule 245 days ago \| hide \| past \| pdf \| discuss
1634.	Weird Generalization and Inductive Backdoors: New Ways to Corrupt LLMs (arxiv.org)
	3 points by PaulHoule 245 days ago \| hide \| past \| pdf \| discuss
1635.	The SWE-Bench Illusion: When LLMs Remember Instead of Reason (arxiv.org)
	2 points by cadabrabra 245 days ago \| hide \| past \| pdf \| discuss
1636.	YuriiFormer: A Suite of Nesterov-Accelerated Transformers (arxiv.org)
	2 points by kelseyfrog 245 days ago \| hide \| past \| pdf \| discuss
1637.	3DGS-Drag: Dragging Gaussians for Intuitive Point-Based 3D Editing (arxiv.org)
	3 points by PaulHoule 245 days ago \| hide \| past \| pdf \| discuss
1638.	Semi-Autonomous Mathematics Discovery with Gemini: Erdős Problems Case Study (arxiv.org)
	1 point by tzury 245 days ago \| hide \| past \| pdf \| discuss
1639.	Forcing and Diagnosing Failure Modes of Fourier Neural Operators (arxiv.org)
	3 points by TimorousBestie 246 days ago \| hide \| past \| pdf \| 1 comment
1640.	Teaching Models to Teach Themselves: Reasoning at the Edge of Learnability (arxiv.org)
	4 points by gmays 246 days ago \| hide \| past \| pdf \| discuss
1641.	Demystifying ARM SME to Optimize General Matrix Multiplications (arxiv.org)
	88 points by matt_d 247 days ago \| hide \| past \| pdf \| 19 comments
1642.	VaultGemma: A Differentially Private LLM (arxiv.org)
	1 point by PaulHoule 247 days ago \| hide \| past \| pdf \| discuss
1643.	Magellan: Autonomous Discovery of Compiler Optimization Heuristics w/AlphaEvolve (arxiv.org)
	4 points by matt_d 247 days ago \| hide \| past \| pdf \| discuss
1644.	Proc3D: Procedural 3D Generation and Parametric Editing of 3D Shapes with LLMs (arxiv.org)
	5 points by PaulHoule 247 days ago \| hide \| past \| pdf \| discuss
1645.	Shaping capabilities with token-level data filtering (arxiv.org)
	2 points by brandonb 247 days ago \| hide \| past \| pdf \| 1 comment
1646.	Self-Distillation Enables Continual Learning (arxiv.org)
	2 points by simonpure 247 days ago \| hide \| past \| pdf \| discuss
1647.	The End of Transformers (2025) (arxiv.org)
	1 point by teleforce 247 days ago \| hide \| past \| pdf \| discuss
1648.	Qwen3-ASR Technical Report (arxiv.org)
	7 points by _____k 247 days ago \| hide \| past \| pdf \| discuss
1649.	Lost in the Middle: How Language Models Use Long Contexts (2023) (arxiv.org)
	2 points by wslh 248 days ago \| hide \| past \| pdf \| discuss
1650.	Scaling Embeddings Outperforms Scaling Experts in Language Models (arxiv.org)
	1 point by simonpure 248 days ago \| hide \| past \| pdf \| discuss
	More