limoce's submissions | Hacker News

1.		Cursor's 'Composer 2' model is apparently just Kimi K2.5 with RL fine-tuning (reddit.com)
		1 point by limoce 1 day ago \| past \| 1 comment
2.		Add NPU support for AICore for Gemma4 model (github.com/google-ai-edge)
		2 points by limoce 12 days ago \| past \| discuss
3.		SteptronOss: Lightweight, AI-native training framework for large language models (github.com/stepfun-ai)
		1 point by limoce 18 days ago \| past
4.		Adding Support for Qwen3.5 (github.com/huggingface)
		5 points by limoce 41 days ago \| past
5.		Reduce Complexity of Matmul to O(N^2logN) by Asymptotically Optimal Quantum Algo (arxiv.org)
		4 points by limoce 43 days ago \| past \| 1 comment
6.		Learning from context is harder than we thought (tencent.com)
		245 points by limoce 46 days ago \| past \| 136 comments
7.		Residual Context Diffusion (yuezhouhu.github.io)
		1 point by limoce 48 days ago \| past
8.		Step 3.5 Flash (stepfun.com)
		7 points by limoce 48 days ago \| past \| 1 comment
9.		LongCat-Flash-Lite 100B A3B Technical Report [pdf] (huggingface.co)
		2 points by limoce 51 days ago \| past
10.		Jinja.cpp: A single-header C++11 Jinja2 template engine for LLM chat templates (github.com/wangzhaode)
		1 point by limoce 3 months ago \| past
11.		Mini-SGLang: A lightweight yet high-performance inference framework for LLM (github.com/sgl-project)
		2 points by limoce 3 months ago \| past
12.		Google drops Gemini 3 Pro image preview (reddit.com)
		2 points by limoce 4 months ago \| past
13.		DGX Spark may have only half the performance claimed (reddit.com)
		2 points by limoce 4 months ago \| past
14.		You Can Cool Chips with Lasers (ieee.org)
		8 points by limoce 4 months ago \| past
15.		Fully Integrated 2D Flash Chip Unveiled (bioengineer.org)
		2 points by limoce 5 months ago \| past
16.		Linux 6.18 UDP receive performance improved by 47%, under DDoS (kernel.org)
		33 points by limoce 5 months ago \| past \| 2 comments
17.		OpenAI 2025 ICPC Submissions (github.com/openai)
		1 point by limoce 5 months ago \| past
18.		Unveiling Silicon Art: Dieshots of Microchip Masterpieces (dieshot.com)
		3 points by limoce 6 months ago \| past
19.		AetherCode: Evaluating LLMs' Ability to Win in Premier Programming Competitions (arxiv.org)
		1 point by limoce 6 months ago \| past
20.		Operation Costs in CPU Clock Cycles (2016) (ithare.com)
		32 points by limoce 7 months ago \| past \| 8 comments
21.		Principles and Methodologies for Serial Performance Optimization (usenix.org)
		1 point by limoce 7 months ago \| past
22.		The Koala Benchmarks for the Shell (kben.sh)
		2 points by limoce 7 months ago \| past
23.		SmallThinker: A Family of Efficient LLMs Natively Trained for Local Deployment (arxiv.org)
		2 points by limoce 7 months ago \| past
24.		Step3 Technical Report [pdf] (github.com/stepfun-ai)
		1 point by limoce 7 months ago \| past
25.		FP8 is ~100 tflops faster when the kernel name has "cutlass" in it (twitter.com/cis_female)
		243 points by limoce 8 months ago \| past \| 107 comments
26.		Polaris: A Post-training recipe for scaling RL on Advanced Reasoning models (hkunlp.github.io)
		4 points by limoce 8 months ago \| past \| 1 comment
27.		Overclocking LLM Reasoning: Monitoring and Controlling LLM Thinking Path Lengths (royeisen.github.io)
		63 points by limoce 8 months ago \| past
28.		Neutrino: Probing-Based eBPF-Like GPU Kernel Profiling (github.com/open-neutrino)
		2 points by limoce 8 months ago \| past
29.		Machine Learning Conferences Should Establish "Refutations and Critiques" Track (arxiv.org)
		3 points by limoce 8 months ago \| past
30.		SuperGPQA: Scaling LLM Evaluation Across 285 Graduate Disciplines (supergpqa.github.io)
		2 points by limoce on March 4, 2025 \| past
		More