New Links | ML News

ML Newsnew | past | best | rss | submit

721.	Position: Coding Benchmarks Are Misaligned with Agentic Software Engineering (arxiv.org)
	2 points by wek 99 days ago \| hide \| past \| pdf \| discuss
722.	Tapered Language Models (arxiv.org)
	2 points by sonabinu 99 days ago \| hide \| past \| pdf \| discuss
723.	Red-Teaming the Agentic Red-Team (arxiv.org)
	3 points by infwhispers 100 days ago \| hide \| past \| pdf \| discuss
724.	Combining LLMs Rarely Beats the Best Single Model, I tested 67 frontier models (arxiv.org)
	1 point by josefchen 100 days ago \| hide \| past \| pdf \| discuss
725.	Mapping Networks: CVPR 2026 Best Paper Award Nominee (arxiv.org)
	4 points by aurenvale 100 days ago \| hide \| past \| pdf \| 1 comment
726.	A Structured Generation Framework for Transforming Scientific Papers into Patent (arxiv.org)
	2 points by teleforce 100 days ago \| hide \| past \| pdf \| discuss
727.	Position: Stop Anthropomorphizing Intermediate Tokens as Reasoning/Thinking Trac (arxiv.org)
	4 points by xiaoyu2006 100 days ago \| hide \| past \| pdf \| discuss
728.	Reading AI Model Compilation in MLIR Through the Lens of Formal Theories (arxiv.org)
	2 points by matt_d 100 days ago \| hide \| past \| pdf \| discuss
729.	David vs. Goliath in Next Activity Prediction: Argmax vs. LSTM, Transformer, LLM (arxiv.org)
	2 points by hramezani 101 days ago \| hide \| past \| pdf \| discuss
730.	Autodata: An agentic data scientist to create high quality synthetic data (arxiv.org)
	4 points by root-parent 101 days ago \| hide \| past \| pdf \| 1 comment
731.	Wikipedia advocacy shapes LLM values (arxiv.org)
	3 points by 50kIters 101 days ago \| hide \| past \| pdf \| discuss
732.	The False Promise of Imitating Proprietary LLMs (arxiv.org)
	2 points by handfuloflight 101 days ago \| hide \| past \| pdf \| discuss
733.	IatroBench: Pre-Registered Evidence of Iatrogenic Harm from AI Safety Measures (arxiv.org)
	3 points by NavinF 101 days ago \| hide \| past \| pdf \| discuss
734.	World Action Models: A Survey (arxiv.org)
	4 points by simonpure 101 days ago \| hide \| past \| pdf \| discuss
735.	Code as Agent Harness (arxiv.org)
	5 points by matt_d 101 days ago \| hide \| past \| pdf \| 1 comment
736.	LLMs use "safety" specific neuron layers to identify vulnerabilities in code (arxiv.org)
	5 points by summarity 101 days ago \| hide \| past \| pdf \| 3 comments
737.	Submodular Context Selection as a Pluggable Engine for LLM Agents (arxiv.org)
	2 points by Elof 102 days ago \| hide \| past \| pdf \| discuss
738.	DualPath: Breaking the Storage Bandwidth Bottleneck in Agentic LLM Inference (arxiv.org)
	2 points by yogthos 102 days ago \| hide \| past \| pdf \| discuss
739.	The Promptware Kill Chain (arxiv.org)
	3 points by wslh 102 days ago \| hide \| past \| pdf \| discuss
740.	Qwen-AgentWorld: Language World Models for General Agents (arxiv.org)
	199 points by ilreb 102 days ago \| hide \| past \| pdf \| 55 comments
741.	Lift4D: Harmonizing Single-View 3D Estimation for 4D Reconstruction In-the-Wild (arxiv.org)
	3 points by MediaSquirrel 103 days ago \| hide \| past \| pdf \| discuss
742.	Inference Compute Shapes Frontier LLM Evaluation (arxiv.org)
	2 points by matt_d 103 days ago \| hide \| past \| pdf \| discuss
743.	Concordia: JIT-Compiled Persistent-Kernel Checkpt for Fault-Tolerant Inference (arxiv.org)
	2 points by matt_d 103 days ago \| hide \| past \| pdf \| discuss
744.	Confidence estimation is a better metric than agreement for LLM judges (arxiv.org)
	3 points by rapiddev 103 days ago \| hide \| past \| pdf \| discuss
745.	PsychoPass: Geometric Profiling of Multi-Turn Adversarial LLM Conversations (arxiv.org)
	3 points by Anon84 103 days ago \| hide \| past \| pdf \| discuss
746.	The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A" (2023) (arxiv.org)
	25 points by Anon84 103 days ago \| hide \| past \| pdf \| 46 comments
747.	Unlimited OCR Works (arxiv.org)
	3 points by ilreb 103 days ago \| hide \| past \| pdf \| discuss
748.	Tapered Language Models (arxiv.org)
	3 points by E-Reverance 103 days ago \| hide \| past \| pdf \| discuss
749.	Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models (arxiv.org)
	55 points by teleforce 103 days ago \| hide \| past \| pdf \| 9 comments
750.	VibeThinker: 3B param model that beats Opus 4.5 on reasoning with novel SFT+GRPO (arxiv.org)
	398 points by timhigins 103 days ago \| hide \| past \| pdf \| 205 comments
	More