New Links | ML News

ML Newsnew | past | best | rss | submit

1231.	Foundations of Polar Linear Algebra (arxiv.org)
	3 points by znpy 180 days ago \| hide \| past \| pdf \| discuss
1232.	Frequent ChatGPT users are accurate detectors of AI-generated text (2025) (arxiv.org)
	11 points by croemer 180 days ago \| hide \| past \| pdf \| 2 comments
1233.	SlopCodeBench: Benchmarking How Coding Agents Degrade over Long-Horizon Task (arxiv.org)
	1 point by mohsen1 181 days ago \| hide \| past \| pdf \| discuss
1234.	Show HN: A Framework for Evaluating Coding Agents on Sequential SWE (arxiv.org)
	1 point by tdchaitanya 181 days ago \| hide \| past \| pdf \| discuss
1235.	Attention Residuals (arxiv.org)
	2 points by djhemath 181 days ago \| hide \| past \| pdf \| 1 comment
1236.	Brevity Constraints Reverse Performance Hierarchies in Language Models (arxiv.org)
	1 point by handfuloflight 181 days ago \| hide \| past \| pdf \| discuss
1237.	Test-Time Scaling Makes Overtraining Compute-Optimal (arxiv.org)
	1 point by matt_d 181 days ago \| hide \| past \| pdf \| discuss
1238.	Analyzing Reverse Address Translation Overheads in Multi-GPU Scale-Up Pods (arxiv.org)
	1 point by matt_d 181 days ago \| hide \| past \| pdf \| discuss
1239.	Optimizing Time, Cost, and Generalization in Distributed Large-Batch Training (arxiv.org)
	2 points by PaulHoule 181 days ago \| hide \| past \| pdf \| discuss
1240.	Model2Kernel: Model-Aware Symbolic Execution for Safe CUDA Kernels (arxiv.org)
	2 points by PaulHoule 181 days ago \| hide \| past \| pdf \| discuss
1241.	Provider-Dependent Energy Effects of Prompt Compression (arxiv.org)
	1 point by PaulHoule 181 days ago \| hide \| past \| pdf \| discuss
1242.	Benchmark-Dependent Output Dynamics in LLM Prompt Compression (arxiv.org)
	1 point by PaulHoule 181 days ago \| hide \| past \| pdf \| discuss
1243.	Show HN: WebGPU LLM inference comprehensive benchmark (arxiv.org)
	2 points by yu3zhou4 182 days ago \| hide \| past \| pdf \| 2 comments
1244.	TELeR: A General Taxonomy of LLM Prompts for Benchmarking Complex Tasks (2023) (arxiv.org)
	2 points by locknitpicker 183 days ago \| hide \| past \| pdf \| discuss
1245.	Meta-Harness: End-to-End Optimization of Model Harnesses (arxiv.org)
	3 points by kstonekuan 183 days ago \| hide \| past \| pdf \| discuss
1246.	Show HN: Signals – finding the most informative agent traces without LLM judges (arxiv.org)
	3 points by sparacha 183 days ago \| hide \| past \| pdf \| discuss
1247.	GTO Wizard Benchmark (arxiv.org)
	1 point by EvgeniyZh 183 days ago \| hide \| past \| pdf \| discuss
1248.	R2R: Navigating Divergent Reasoning with Small-Large Model Token Routing (2025) (arxiv.org)
	2 points by davikr 183 days ago \| hide \| past \| pdf \| discuss
1249.	Embarrassingly simple self-distillation improves code generation (arxiv.org)
	658 points by Anon84 183 days ago \| hide \| past \| pdf \| 201 comments
1250.	Reasoning models encode tool choices before they start reasoning (arxiv.org)
	3 points by diwank 184 days ago \| hide \| past \| pdf \| discuss
1251.	Void: Video Object and Interaction Deletion (arxiv.org)
	2 points by jonbaer 184 days ago \| hide \| past \| pdf \| discuss
1252.	Nvidia Nemotron 3: Efficient and Open Intelligence (arxiv.org)
	1 point by andsoitis 184 days ago \| hide \| past \| pdf \| discuss
1253.	Do All Languages Cost the Same? Tokenization in the Era of Commercial LLMs (arxiv.org)
	2 points by Anon84 184 days ago \| hide \| past \| pdf \| discuss
1254.	PolarQuant: Quantizing KV Caches with Polar Transformation (arxiv.org)
	3 points by felineflock 184 days ago \| hide \| past \| pdf \| discuss
1255.	Embarrassingly Simple Self-Distillation Improves Code Generation (arxiv.org)
	4 points by jryio 184 days ago \| hide \| past \| pdf \| 3 comments
1256.	Latent Introspection: Models Can Detect Prior Concept Injections (arxiv.org)
	2 points by tosh 185 days ago \| hide \| past \| pdf \| discuss
1257.	When a reasoning LLM chooses, which comes first: thought or decision? (arxiv.org)
	1 point by dotvignesh 185 days ago \| hide \| past \| pdf \| discuss
1258.	Alignment Whack-a-Mole (arxiv.org)
	2 points by ai_critic 185 days ago \| hide \| past \| pdf \| discuss
1259.	Composer 2 Technical Report (arxiv.org)
	1 point by gmays 185 days ago \| hide \| past \| pdf \| discuss
1260.	Bio-inspired LLM reasoning architecture with mycological dormancy (arxiv.org)
	2 points by odi_s 185 days ago \| hide \| past \| pdf \| discuss
	More