News
AI, cybersecurity, machine learning, scraping & programming news.
LLMs Are Turing-Complete
Transformers Can Compute Anything Commentators often claim that large language models will always fail at certain tasks because of fundamental architectural limits. Those claims are wrong. Transformer models are Turing-complete. T...
LeanLLM: An LLM SDK for applications written in Lean 4
A single Lean 4 client for every major LLM provider theoriclabs has published leanllm, a Lean 4 library that wraps Anthropic Claude, OpenAI, Google Gemini, DeepSeek, Kimi/Moonshot, Z.AI GLM, Qwen, MiniMax, xAI Grok, and Mistral be...
ChunkAct: Action Chunking for Agentic Robotics
Agentic robotics places a large language model (LLM) agent in the robot control loop, where it composes perception, motion, and vision-language-action (VLA) tools under the guidance of a memory and thereby generalizes to new tasks...
Dog-Breed Text-to-image Flow Matching from Pretraining to Fine-Tuning on 4 Consumer GPUs
We generally perceive that training text-to-image diffusion models requires heavy compute: it is true if we aim at a comprehensive frontier model. This understanding prevents people from training their own models when getting star...
Tev: Scaling Long-Horizon Agentic Environments with Dense Progress Rewards
The problem this paper tackles is the failure of open‑source models to operate autonomously on long‑horizon tasks that span economically valuable work, from multi‑hour computations to multi‑day projects. Such a capability is essen...
PixelMoE: Scaling Pixel-Space Diffusion with LLM-Style Mixture-of-Experts
Pixel-space diffusion models have long appealed as an alternative to latent-space approaches, offering direct modeling of raw images without dependence on a learned tokenizer. Yet scaling these models has followed dense strategies...
Two Clocks: Why Larger Rollout Batches Learn Slower but Fail on Schedule in Reinforcement Learning with Verifiable Rewards
Reinforcement learning with verifiable rewards RLVR relies on costly model-generated responses, motivating larger rollout batches and square-root learning-rate scaling. Reported to approximately preserve progress per sample with o...
Show HN: Jevcrawl – A web crawler for AI agents with no LLM in the crawl loop
A new MCP server and CLI called jevcrawl gives AI agents a way to crawl websites that finds the answer page more often, costs a fraction of current approaches, and returns only the passages that matter. The system uses a small dec...
Show HN: Deslop – Oxlint rules for AI-generated TypeScript
Oxlint has acquired a new set of opinionated rules designed to reject low-evidence, low-signal code patterns increasingly produced by AI coding assistants. The project, deslop, ships as the npm package ai-deslop and provides a cur...
FrontierML: A challenging, end-to-end machine learning benchmark
Scaling capital access across agent runs FrontierML enters the landscape as a benchmark from Induction Labs that measures how effectively agents convert budget into model performance on end-to-end machine learning tasks. The proje...