Action as Choice Expressed Through Movement Toward a Goal: a Framework for Overcoming Inaction
I’ve reached a point in my life where I realize that everything I want sits on the other side of action. Earlier this week, I randomly…
Latest
A software engineering interview question I like: computing the median
Article URL: https://krisshamloo.com/blog/007 Comments URL: https://news.ycombinator.com/item?id=48839434 Points: 25 # Comments: 11
Remote Attestation
Article URL: https://www.liamcvw.com/p/remote-attestation Comments URL: https://news.ycombinator.com/item?id=48839397 Points: 6 # Comments: 1
the end is nigh
the end is nighdoomslide: But the goat chosen by lot as the scapegoat shall be presented alive before the Lord to be used for making atonement…
MIRA: Multiplayer Interactive World Models Trained on Rocket League
https://mira-wm.com/blog-post/ Comments URL: https://news.ycombinator.com/item?id=48839355 Points: 3 # Comments: 0
Can you explain the concept behind each of the main size ranges of LLM models, as in, what hardware setups the different size niches are meant to fit into (~30b, ~70b, ~120b, ~230b, etc). Like is it mainly based on pro hardware sizing for 8-bit, or consumer GPU vram for ~Q4, or some mixture?
I am curious about intended sizings of the main size niches of the popular local LLM models. As in, we can see there is a major…
RT Alfredo Canziani @ ICML: Come and learn about «Anatomy of Massive Activations and Attention Sinks», 📆 today (Thu) at 🕥 10:30, 📍 HALL A #…
RT Alfredo Canziani @ ICMLCome and learn about «Anatomy of Massive Activations and Attention Sinks», 📆 today (Thu) at 🕥 10:30, 📍 HALL A #3002. #ICML2026
b9934
ggml-webgpu: tune subgroup split (d_split) in flash_attn_vec (#25418) macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramework Linux:…
Robbyant Releases LingBot-VLA 2.0: An Open-Source 6B Vision-Language-Action (VLA) Model for Cross-Embodiment Robot Manipulation
Ant Group's Robbyant has released LingBot-VLA 2.0, an Apache-2.0 vision-language-action model for cross-embodiment robot manipulation. The 6B checkpoint is pretrained on roughly 60,000 hours of data,…
Grok speaks like a robot. Actually like the @grok doing responses here. Terse, stilted, enumerating options. ESL vibes. Agent Grok I like it. totally …
Grok speaks like a robot. Actually like the @grok doing responses here. Terse, stilted, enumerating options. ESL vibes. Agent GrokI like it. totally useless for many…
Modular Pretraining Enables Access Control
Full author list: Ethan Roland*, Murat Cubuktepe*, Erick Martinez*, Stijn Servaes, Keenan Pepper, Mike Vaiana, Diogo Schwerz de Lucena, Judd Rosenblatt, Addie Foote, Cem Anil, Alex…
Qwen3.5 122B is the best?
I’m using Opencode and a computer with 128gb. So maybe the results would be different on system. I’ve exhaustingly tried Qwen3.6 27B and Qwen3.6 33B. I…
LiteRT.js, Google's high performance Web AI Inference
We're excited to introduce LiteRT.js, the newest member of the LiteRT family! LiteRT.js is our powerful solution for running machine learning models directly in the browser,…
Unmasking On-Policy Distillation: Where It Helps, Where It Hurts, and Why
On-policy distillation offers dense, per-token supervision for training reasoning models; however, it remains unclear under which conditions this signal is beneficial and under which it is…
Recursive Language Models Meet Uncertainty: The Surprising Effectiveness of Self-Reflective Program Search for Long Context
Long-context handling remains a core challenge for language models: even with extended context windows, models often fail to reliably extract, reason over, and use the information…
Ben Bernanke appointed to Anthropic’s Long-Term Benefit Trust
Ben Bernanke appointed to Anthropic’s Long-Term Benefit Trust
Incentivizing Temporal-Awareness in Egocentric Video Understanding Models
Multimodal large language models (MLLMs) have recently shown strong performance in visual understanding, yet they often lack temporal awareness, particularly in egocentric settings where reasoning depends…
Ollama: all aboard open models
Serving 8.9 million developers, Ollama has raised $88M from Benchmark, Theory Ventures, 8VC, and many incredible angel investors.
Introducing a way to reflect on how you use Claude
Introducing a way to reflect on how you use Claude
Seems to suggest that if it stops being the case that there's 3 labs which are all roughly equally good, competing each others margins away, the provi…
Seems to suggest that if it stops being the case that there's 3 labs which are all roughly equally good, competing each others margins away, the…
What GUI-first coding tool tool are you pairing your local LLMs with? Opencode isn’t it for me.
I've grown very frustrated with OpenCode. The web GUI and desktop app ideas are good, but the execution not so much. The GUI is lacking so…
There Should Be More AI Safety Hubs
Right now, most AI safety talent is concentrated in just a few cities: San Francisco, Berkeley, London, Oxford, and Washington D.C. Some other cities also deserve…
Rewriting Bun in Rust
Rewriting Bun in Rust Jarred Sumner has been promising this blog post (since May 9th) about his Zig to Rust rewrite of Bun for significantly longer…
We’re pleased to have collaborated with AE Studio on this research. Read more here: https://www.anthropic.com/research/off-switch-dual-use
We’re pleased to have collaborated with AE Studio on this research. Read more here: https://www.anthropic.com/research/off-switch-dual-useAE Studio: New research! Some AI capabilities are both helpful and dangerous.…