New Links | ML News

ML Newsnew | past | best | rss | submit

511.	Every Time I Hire a Linguist, Inference Costs Go Down (arxiv.org)
	3 points by cwbuilds 66 days ago \| hide \| past \| pdf \| discuss
512.	CryptanalysisBench: Can LLMs Do Cryptanalysis? (arxiv.org)
	1 point by zdw 66 days ago \| hide \| past \| pdf \| discuss
513.	Teaching agents to predict and pre-execute their next tool call (arxiv.org)
	6 points by rotariuvladimir 67 days ago \| hide \| past \| pdf \| discuss
514.	Messier: A High-Resolution Corpus for Cross-Benchmark Agent Evaluation (arxiv.org)
	3 points by root-parent 67 days ago \| hide \| past \| pdf \| discuss
515.	Visual prompt engineering for video models (arxiv.org)
	4 points by root-parent 67 days ago \| hide \| past \| pdf \| discuss
516.	Handbook.md shows that long policy documents do not reliably govern agents (arxiv.org)
	325 points by spIrr 67 days ago \| hide \| past \| pdf \| 209 comments
517.	Detecting CSAM Text-to-Image LoRAs from Weights (arxiv.org)
	6 points by sbulaev 67 days ago \| hide \| past \| pdf \| discuss
518.	Do LLMs Know What Other LLMs Don't? Peer-Probing as Scalable Evaluation (arxiv.org)
	4 points by tcp_handshaker 67 days ago \| hide \| past \| pdf \| discuss
519.	Adaptive Agentic Attacks on LLM Vulnerability Detectors via Adversarial Comments (arxiv.org)
	3 points by tcp_handshaker 67 days ago \| hide \| past \| pdf \| discuss
520.	ProofCouncil: An LLM Agent for Solving Open Mathematical Problems (arxiv.org)
	3 points by frozenseven 67 days ago \| hide \| past \| pdf \| 1 comment
521.	Mapping CVEs to Mitre ATT&CK Techniques (arxiv.org)
	3 points by adulau 67 days ago \| hide \| past \| pdf \| 1 comment
522.	Certified in Theory, Broken in Practice: Assumption Gaps in Cryptographic Model (arxiv.org)
	3 points by sbulaev 67 days ago \| hide \| past \| pdf \| discuss
523.	The SpiNNaker2 chip: a many-core platform for brain-inspired computing (arxiv.org)
	1 point by Jimmc414 67 days ago \| hide \| past \| pdf \| discuss
524.	Distilling proprietary model reasoning into open-source search agents (arxiv.org)
	5 points by cpard 67 days ago \| hide \| past \| pdf \| discuss
525.	CryptanalysisBench: Can LLMs Do Cryptanalysis? (arxiv.org)
	2 points by rvz 67 days ago \| hide \| past \| pdf \| discuss
526.	"Uncensored" open LLMs are measurably more optimistic than their base models (arxiv.org)
	43 points by oleczek 67 days ago \| hide \| past \| pdf \| 22 comments
527.	Agentic Permissions Policy Algebra for Taint Confinement in LLM Agents (arxiv.org)
	2 points by zhinit 67 days ago \| hide \| past \| pdf \| discuss
528.	Kimi Linear: An Expressive, Efficient Attention Architecture (2025) (arxiv.org)
	306 points by ronfriedhaber 68 days ago \| hide \| past \| pdf \| 133 comments
529.	LoopLynx: A Scalable Dataflow Architecture for Efficient LLM Inference (arxiv.org)
	2 points by binyu 68 days ago \| hide \| past \| pdf \| discuss
530.	Tag Questions and the Generational Reversal of Sycophancy Across 45 Language (arxiv.org)
	2 points by sbulaev 68 days ago \| hide \| past \| pdf \| discuss
531.	Neural Representation of Minimal Surfaces (arxiv.org)
	2 points by E-Reverance 68 days ago \| hide \| past \| pdf \| discuss
532.	Show HN: A 6M-token movable window on a single 46GB GPU (arxiv.org)
	7 points by Wetime 68 days ago \| hide \| past \| pdf \| 16 comments
533.	Protocol-Level Attacks on Agentic Commerce Platforms: Taxonomy and Defense (arxiv.org)
	2 points by sbulaev 68 days ago \| hide \| past \| pdf \| discuss
534.	Lost in Context: Addressing Context Anxiety in Large Language Models (arxiv.org)
	2 points by StatsAreFun 68 days ago \| hide \| past \| pdf \| discuss
535.	The Polynomial-Time Low-Degree Conjecture Is False (arxiv.org)
	4 points by MarcoDewey 68 days ago \| hide \| past \| pdf \| discuss
536.	Hardware-Software Co-Design for Float16 On-Device Training on RISC-V Single-Core (arxiv.org)
	2 points by Jimmc414 68 days ago \| hide \| past \| pdf \| discuss
537.	Paper – Agent Memory (arxiv.org)
	2 points by qspencer 69 days ago \| hide \| past \| pdf \| discuss
538.	Every Model Cheats: Prompt-Level Mitigation of Cheating on Offensive Cyber Tasks (arxiv.org)
	1 point by sbulaev 69 days ago \| hide \| past \| pdf \| discuss
539.	Do Agent Benchmarks Measure Capability? Protocol Validity in the Age of Agentic (arxiv.org)
	2 points by sbulaev 69 days ago \| hide \| past \| pdf \| discuss
540.	Frontier LLMs drop from 83% to 43% once reasoning has to chain across domains (arxiv.org)
	2 points by MarcoDewey 69 days ago \| hide \| past \| pdf \| 1 comment
	More