c/technology

Amazon’s recent pivot on the Selling Partner API is an encouraging reminder that developer pushback can breach walled gardens.

Vancouver Police Department Partnering with our community for excellence and innovation in public safety. Partnering with our community for excellence and innovation in public safety. Learn More Got a question about our services? Call us : 1.800.123.4567 VPD News Nothing Found

An AI system that solves open mathematics problems using Lean 4 — every solution kernel-verified and downloadable.

A PJM market watchdog calls the shift a "massive wealth transfer" to tech companies, but there's a peak-demand loophole.

As heatwaves become more frequent across Europe, Switzerland is confronting an unfamiliar question: whether rules around air conditioning should become less restrictive. Air Conditioning © Tang9024

Buildkite is a CI developers can love, with a hybrid approach that gives a front seat to the increased velocity of software development in 2026.

rob/rob.mw on remark.ing

agents are the next generation of users of the internet, but the web keeps turning them away. not anymore. today we're launching tilion, for enabling the next evolution of the web. sign up for closed beta today.

Written pieces, talks, and other bits by Zach Holman.

Scott

Japanese leaders have privately approached allies such as the United States, Australia and Germany in recent months for advice on technology, staffing and priorities.

Scaling pre-training, post-training, and test-time compute have become the central paradigms for improving the capabilities of LLMs. In this work, we identify verification, the ability to determine the correctness of a solution, as a new scaling axis. To unlock this and demonstrate its effectiveness, we introduce LLM-as-a-Verifier, a general-purpose verification framework that provides fine-grained feedback for agentic tasks without requiring additional training. Unlike standard LM judges that prompt LLMs to produce discrete scores for candidate solutions, LLM-as-a-Verifier computes the expectation over the distribution of scoring token logits to generate continuous scores. This probabilistic formulation enables verification to scale along multiple dimensions: (1) score granularity, (2) repeated evaluation, and (3) criteria decomposition. In particular, we show that scaling the scoring granularity leads to better separation between positive and negative solutions, resulting in more calibrated comparisons. Moreover, scaling repeated evaluation and criteria decomposition consistently lead to additional gains in verification accuracy through variance and complexity reduction. We further introduce a cost-efficient ranking algorithm for selecting the best solution among candidates using the verifier's continuous scores. LLM-as-a-Verifier achieves state-of-the-art performance on Terminal-Bench V2 (86.5%), SWE-Bench Verified (78.2%), RoboRewardBench (87.4%), and MedAgentBench (73.3%). Beyond verification, the fine-grained signals from LLM-as-a-Verifier can also serve as a proxy for estimating task progress. We build an extension for Claude Code, enabling developers to monitor and improve their own agentic systems. Finally, we show that LLM-as-a-Verifier can provide dense feedback for RL, improving the sample efficiency of SAC and GRPO on robotics and mathematical reasoning benchmarks.

ARDY is an autoregressive diffusion model for interactive human motion generation with online text prompting and flexible kinematic constraints.

Field notes on shipping fast, privacy-first, evergreen web products.