Can't Find Benchmark and download link ? See here 👇https://huggingface.co/AngelSlim/Hy3-GGUFTencent Hy: We’ve just released the 1-bit & 4-bit version of Hy3, a flagship-scale 295B model that…
RT Michael NicollsCongrats to the @Starlink team on the release of the new Starlink V5 terminal, designed with high performance, reliability, and scalability in mind based…
opencl: do not use clCreateBufferWithProperties when targeting CL 2.x (#25673) macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS…
In this paper, we propose SpectraReward, a training-free reward function that turns pretrained MLLMs into off-the-shelf reward models for image-generation reinforcement learning. Instead of asking the…
Reading one of the viral "anti-slop" markdown files for agents and realizing it is basically a 100% AI written and will basically hyperstition your outputs into…
I endorse this bit. Graham died and instantly went to Republican Heaven, now he'll be venerated like John McCain. Patriot, builder, stout defender of our common…
Article URL: https://fabiensanglard.net/jurrasic_park_computers/index.html Comments URL: https://news.ycombinator.com/item?id=48915709 Points: 16 # Comments: 2
How many AI safety papers are at the big ML conferences, what do they study, and who writes them? A comprehensive analysis.> Website: https://ai-safety-tracker-website.vercel.app/> Data, code…
That Hy3 test from Jen @jenzhuscott was really impressive. Thank you. 🥰Jen Zhu: Been power testing @TencentHunyuan Hy3 & not surprised why it’s the most used…
I spent the last month running a persistent personal-AI system on top of Claude Code. This is a field note, not a benchmark, and the hypothesis…
Grok 4.5 is worth tryingX Freeze: Grok 4.5 just ranked #2 on the FrontierSWE leaderboard, outperforming Claude Opus 4.8 and GPT-5.5It delivers top-tier coding performance with…
Most "run a model in your browser" demos still lean on a server or WebGPU. This one doesn't: it's a pure WebAssembly inference engine for LiquidAI's…
Order the Starlink V5 terminal!Starlink: The next generation Starlink Kit is designed to deliver reliable, high-speed home internet. Starlink V5 has a smaller form factor and…
this is just coolDerrick Choi: Try out the Visualize plugin (in preview) in Codex. Some ideas click faster when you can see and interact with it.Here’s…
LLM-based coding agents have significantly advanced automated software issue resolution, yet they remain highly prone to factual errors caused by insufficient repository understanding. Recent methods attempt…
Title says it, is anyone having any luck with the Ternary Bonsai 27B DFlash? I have been playing around with it but have seen no speed…
Article URL: https://worldcrunch.com/focus/green-or-gone/global-warming-at-3c-by-2050-what-s-behind-the-new-german-climate-warning/ Comments URL: https://news.ycombinator.com/item?id=48915244 Points: 10 # Comments: 2
RT Yun-Ta TsaiPeople may disagree on which frontier model is best, but they all seem to agree the current X algorithm is superb.
with chatgpt work & sol, i'm finding it incredibly joyful to just ask any question about the business and have it be thoroughly researched and answered.realizing…
RT Bhavin Jawade𝗗𝗲𝗲𝗽𝗦𝗲𝗲𝗸 𝗩𝟰 𝗽𝗼𝘀𝘁-𝘁𝗿𝗮𝗶𝗻𝗶𝗻𝗴 𝗿𝗲𝗰𝗶𝗽𝗲DeepSeek V4 came out a couple of months ago. I finally sat down and went through the technical report (https://arxiv.org/abs/2606.19348). In…
Article URL: https://tailscale.com/security-bulletins Comments URL: https://news.ycombinator.com/item?id=48915004 Points: 20 # Comments: 3
I was poking around at a chessformer which mimics human play and made this fun companion app to visualize a lot of the internals of the…
https://techcrunch.com/2026/07/13/satya-nadella-has-issued-a-shocking-warning-to-companies-using-ai/ Venture capitalists have been warning for awhile that OpenAI and Anthropic are getting access to sensitive business information. The risk is the model makers can…
The funding discussions point to investor interest in applying AI to make breakthroughs in life sciences.