New Links | ML News

ML Newsnew | past | best | rss | submit

1.	Process Matters More Than Output for Distinguishing Humans from Machines (arxiv.org)
	1 point by timshell 2 hours ago \| hide \| past \| pdf \| discuss
2.	SoftServe: A Scalable Quasi-Newton Method for Deep Learning (arxiv.org)
	1 point by E-Reverance 20 hours ago \| hide \| past \| pdf \| discuss
3.	PTXBench: Benchmarking and Adapting LLMs for GPU Kernel Optimization (arxiv.org)
	3 points by matt_d 23 hours ago \| hide \| past \| pdf \| discuss
4.	GPU-Initiated Communication: Dissecting Down to the Bone (arxiv.org)
	2 points by matt_d 1 day ago \| hide \| past \| pdf \| discuss
5.	SFT matches RL if you MCMC the training data first (arxiv.org)
	2 points by mrkn1 1 day ago \| hide \| past \| pdf \| discuss
6.	AI Agents Are Vulnerable to Radicalization (arxiv.org)
	3 points by Anon84 1 day ago \| hide \| past \| pdf \| discuss
7.	Fixing GRPO's credit assignment problem without evaluating every step (arxiv.org)
	23 points by mrkn1 1 day ago \| hide \| past \| pdf \| 3 comments
8.	Decoding Looped Transformers Better for Almost Free (arxiv.org)
	1 point by mrkn1 1 day ago \| hide \| past \| pdf \| discuss
9.	Language Drift During RLVR Post-Training (arxiv.org)
	1 point by sbulaev 1 day ago \| hide \| past \| pdf \| discuss
10.	Removing Timing Shortcuts Improves Non-Invasive Brain-to-Text (arxiv.org)
	3 points by sbulaev 1 day ago \| hide \| past \| pdf \| discuss
11.	Scaling Laws for Looped Mixture of Experts (arxiv.org)
	2 points by matt_d 1 day ago \| hide \| past \| pdf \| discuss
12.	Covert Assistance: Helpful LLM Agents Evade Oversight in Multi-Agent Systems (arxiv.org)
	1 point by sbulaev 1 day ago \| hide \| past \| pdf \| 1 comment
13.	Superhuman AI for Stratego (arxiv.org)
	5 points by droidjj 1 day ago \| hide \| past \| pdf \| 1 comment
14.	When Fancy Eviction Fails: Rethinking Cache Replacement for LLM Prefix Reuse (arxiv.org)
	1 point by matt_d 2 days ago \| hide \| past \| pdf \| discuss
15.	Pretraining Latent Information Feedback Transformers with Teacher Supervision (arxiv.org)
	3 points by gmays 2 days ago \| hide \| past \| pdf \| discuss
16.	Decode-Latency Feedback Prefill: A Model-Free Controller (arxiv.org)
	1 point by gauravapiscean 2 days ago \| hide \| past \| pdf \| 1 comment
17.	Thinking Before Thinking: Scaling Agentic Inference Through Meta-Reasoning (arxiv.org)
	2 points by theanonymousone 2 days ago \| hide \| past \| pdf \| discuss
18.	Context Language Models (arxiv.org)
	175 points by emersonmacro 2 days ago \| hide \| past \| pdf \| 51 comments
19.	Learning Steganography Is Easy, Learning Steganographic Reasoning Is Hard (arxiv.org)
	1 point by sbulaev 2 days ago \| hide \| past \| pdf \| discuss
20.	Thinking Before Thinking: Scaling Agentic Inference Through Meta-Reasoning (arxiv.org)
	3 points by matt_d 3 days ago \| hide \| past \| pdf \| discuss
21.	AI as a Compiler: Compiling Triton kernels without the Triton compiler (arxiv.org)
	2 points by matt_d 3 days ago \| hide \| past \| pdf \| discuss
22.	Distillation Defenses Easily Break After Reinforcement Learning (arxiv.org)
	2 points by ollybritton 3 days ago \| hide \| past \| pdf \| discuss
23.	Sycophantic Chatbots Cause Delusional Spiraling, Even in Ideal Bayesians (arxiv.org)
	4 points by luispa 3 days ago \| hide \| past \| pdf \| discuss
24.	Decoupled DiLoCo for Resilient Distributed Pre-Training (arxiv.org)
	1 point by lawrenceyan 3 days ago \| hide \| past \| pdf \| discuss
25.	Purlin: Separating Orchestration from the Datapath of Collectives (arxiv.org)
	3 points by matt_d 3 days ago \| hide \| past \| pdf \| discuss
26.	Context Language Models (arxiv.org)
	6 points by tomatomatomato 3 days ago \| hide \| past \| pdf \| 1 comment
27.	Practical Secrets Extraction Against Black-Box LLMs (arxiv.org)
	1 point by sbulaev 3 days ago \| hide \| past \| pdf \| discuss
28.	Shutdown Sabotage Propensities in Multi- Agent Systems (arxiv.org)
	1 point by baxtr 3 days ago \| hide \| past \| pdf \| discuss
29.	Compiling Triton kernels without the Triton compiler (arxiv.org)
	3 points by 50kIters 3 days ago \| hide \| past \| pdf \| discuss
30.	Jev-as-a-Judge: Accept When Confident, Escalate When Unsure (arxiv.org)
	2 points by nico 3 days ago \| hide \| past \| pdf \| discuss
	More