New Links | ML News

ML Newsnew | past | best | rss | submit

1051.	Visual Generation Unlocks Human-Like Reasoning Through Multimodal World Models (arxiv.org)
	2 points by felineflock 147 days ago \| hide \| past \| pdf \| discuss
1052.	Generative Recommendation for Large-Scale Advertising (arxiv.org)
	2 points by mfiguiere 147 days ago \| hide \| past \| pdf \| discuss
1053.	VDCores: Resource Decoupled Programming and Execution for Asynchronous GPU (arxiv.org)
	2 points by matt_d 147 days ago \| hide \| past \| pdf \| discuss
1054.	Extracting alignment data in open models (arxiv.org)
	2 points by yakkomajuri 147 days ago \| hide \| past \| pdf \| discuss
1055.	Superintelligent Retrieval Agent: The Next Frontier of Information Retrieval (arxiv.org)
	16 points by CharlesW 148 days ago \| hide \| past \| pdf \| discuss
1056.	Towards Compute-Aware In-Switch Computing for LLMs on Multi-GPU Systems (arxiv.org)
	1 point by rbanffy 148 days ago \| hide \| past \| pdf \| discuss
1057.	MoE-Hub Taming Software Complexity for Seamless MoE Overlap on Multi-GPU Systems (arxiv.org)
	1 point by rbanffy 148 days ago \| hide \| past \| pdf \| discuss
1058.	AutoKernel: Autonomous GPU Kernel Optimization via Iterative Agent-Driven Search (arxiv.org)
	4 points by OsamaJaber 148 days ago \| hide \| past \| pdf \| discuss
1059.	ActCam: Zero-Shot Joint Camera and 3D Motion Control for Video Generation (arxiv.org)
	2 points by danborn26 148 days ago \| hide \| past \| pdf \| discuss
1060.	UniPool: A Globally Shared Expert Pool for Mixture-of-Experts (arxiv.org)
	2 points by danborn26 148 days ago \| hide \| past \| pdf \| discuss
1061.	LLMs corrupt your documents when you delegate (arxiv.org)
	479 points by rbanffy 148 days ago \| hide \| past \| pdf \| 201 comments
1062.	LeWorldModel: Stable End-to-End Predictive Architecture from Pixels (arxiv.org)
	6 points by felineflock 148 days ago \| hide \| past \| pdf \| discuss
1063.	IatroBench: Pre-Registered Evidence of Iatrogenic Harm from AI Safety Measures (arxiv.org)
	2 points by NavinF 149 days ago \| hide \| past \| pdf \| 1 comment
1064.	Hallucinations Undermine Trust; Metacognition Is a Way Forward (arxiv.org)
	19 points by gmays 149 days ago \| hide \| past \| pdf \| 7 comments
1065.	Ads in AI Chatbots? An Analysis of How LLMs Navigate Conflicts of Interest (arxiv.org)
	1 point by DeusExMachina 149 days ago \| hide \| past \| pdf \| discuss
1066.	End-to-End Autoregressive Image Generation with 1D Semantic Tokenizer (arxiv.org)
	1 point by gmays 149 days ago \| hide \| past \| pdf \| discuss
1067.	Why Are All LLMs Obsessed with Japanese Culture? (arxiv.org)
	4 points by geox 150 days ago \| hide \| past \| pdf \| 2 comments
1068.	Aesthetic Layout in LLM-Based Slide Generation via Verifiable Rewards (arxiv.org)
	2 points by PaulHoule 150 days ago \| hide \| past \| pdf \| discuss
1069.	Synthetic Computers at Scale for Long-Horizon Productivity Simulation (arxiv.org)
	1 point by gmays 150 days ago \| hide \| past \| pdf \| discuss
1070.	LCM: Lossless Context Management (arxiv.org)
	3 points by Brajeshwar 150 days ago \| hide \| past \| pdf \| discuss
1071.	LAWS: A new transform operation turning LLM inference into cheap cache lookups (arxiv.org)
	7 points by EGreg 151 days ago \| hide \| past \| pdf \| 2 comments
1072.	ProgramBench: Can language models rebuild programs from scratch? (arxiv.org)
	150 points by jonbaer 151 days ago \| hide \| past \| pdf \| 80 comments
1073.	Counting as a minimal probe of language model reliability (arxiv.org)
	4 points by nateb2022 151 days ago \| hide \| past \| pdf \| discuss
1074.	Arcee Trinity Large Technical Report (arxiv.org)
	4 points by tcp_handshaker 151 days ago \| hide \| past \| pdf \| discuss
1075.	Gavel: Towards Rule-Based Safety Through Activation Monitoring (arxiv.org)
	2 points by wslh 151 days ago \| hide \| past \| pdf \| discuss
1076.	Same Voice, Different Lab: On the Homogenization of Frontier LLM Personalities (arxiv.org)
	1 point by Brajeshwar 151 days ago \| hide \| past \| pdf \| discuss
1077.	CommFuse: Hiding Tail Latency via Communication Decomposition and Fusion (arxiv.org)
	5 points by matt_d 152 days ago \| hide \| past \| pdf \| discuss
1078.	Shared Lexical Task Representations Explain Behavioral Variability in LLMs (arxiv.org)
	2 points by PaulHoule 152 days ago \| hide \| past \| pdf \| discuss
1079.	Models hallucinate more than you think (arxiv.org)
	2 points by axelriet 152 days ago \| hide \| past \| pdf \| 1 comment
1080.	Unlocking Long-Context LLM Training via Compiler-Based Sequence Parallelism (arxiv.org)
	2 points by PaulHoule 152 days ago \| hide \| past \| pdf \| discuss
	More