zagwdt's submissions | Hacker News

1.		Score Centering Stabilizes Off-Policy Reinforcement Learning (arxiv.org)
		3 points by zagwdt 7 days ago \| past \| discuss
2.		A/B testing LLMs in production (together.ai)
		2 points by zagwdt 39 days ago \| past
3.		Inference Optimization for MiniMax Sparse Attention (together.ai)
		1 point by zagwdt 3 months ago \| past
4.		DeepSeek V4 in vLLM: Efficient Long-Context Attention (vllm-website-pdzeaspbm-inferact-inc.verc...)
		3 points by zagwdt 5 months ago \| past
5.		Introspective Diffusion Language Models (introspective-diffusion.github.io)
		281 points by zagwdt 5 months ago \| past \| 55 comments
6.		EinsteinArena: Harnessing the collective intelligence of agents in the wild (einsteinarena.com)
		5 points by zagwdt 5 months ago \| past
7.		RL Meets Adaptive Speculative Training (together.ai)
		2 points by zagwdt 5 months ago \| past
8.		Weak models excel at long context tasks (together.ai)
		2 points by zagwdt 6 months ago \| past
9.		TorchSpec: Speculative Decoding Training at Scale (pytorch.org)
		2 points by zagwdt 6 months ago \| past
10.		Flash Attention 4 (together.ai)
		1 point by zagwdt 6 months ago \| past
11.		CoderForge-Preview: SOTA open dataset for training efficient coding agents (together.ai)
		1 point by zagwdt 7 months ago \| past
12.		Two years of vector search at Notion: 10x scale, 1/10th cost (notion.com)
		2 points by zagwdt 7 months ago \| past
13.		Consistency diffusion language models: Up to 14x faster, no quality loss (together.ai)
		219 points by zagwdt 7 months ago \| past \| 96 comments