2026年7月3日 技术热点总结

项目简介：Rent GPUs and run any AI model, paid in SOL. Buy with $RAM and every $RAM spent gets burned.

技术栈：Vue

核心介绍：CA: GTDSxLef3pnLPpChemVDnB6n2CFqoFTMyDecydqGpump AI runs on two things: compute to run it, and models to run. Both are gatekept — GPUs are scarce and paid for with cards through providers that reject half the world, and model access is locked behind per-provider accounts and billing.

项目数据：⭐ 566 Stars，🍴 3 Forks

项目简介：Give your AI agent eyes and hands on iOS Simulator and Android emulator/devices.

技术栈：Swift、accessibility、ai-agents、ai-development、android-emulator、ios-simulator、mobile-automation

核心介绍：Give AI agents the ability to observe and act on iOS Simulator and Android emulator / device screens. App: Settings 402×874 @1 StaticText “Settings” [Content y=120..754] @5 SearchField “Search” @7 Button “Sign in to your iPhone” @9 Button “General” @10 Button “Display & Brightness” @11 Button “Wallpaper”

项目数据：⭐ 459 Stars，🍴 26 Forks

In collaborative dialogue, shared perception does not guarantee shared interpretation. Mutual understanding must be established through interaction. We investigate whether vision-language models (VLMs) can distinguish what could be shared from what has been shared between dialogue participants through grounding. We formulate this as an interpretation-matching task on 13,077 annotated reference expressions from HCRC MapTask dialogues, and evaluate VLMs under systematically controlled manipulations of dialogue context and map-information access. Our results show that providing authentic map i…

Three of the most popular methods for training language models to reason look like three different tricks. They are not. All three adjust a single number: standard deviation, reflecting how much a prompt's sampled answers disagree. When such a model is trained, it answers each problem many times, and an automatic checker marks every answer right or wrong. The standard deviation of those marks measures the disagreement: largest when the answers split evenly between right and wrong, and zero when they all agree. Group Relative Policy Optimization (GRPO) divides by this number, GRPO Done Right…

People overthink; language models over-sample, and the extra effort can talk both into a worse answer. Reasoning systems answer a hard question by sampling it many times (test-time scaling), and the more they draw, the more often a correct answer turns up somewhere, so coverage, the fraction of problems with at least one correct try, climbs and appears to be progress. But a deployed system must return one answer, and choosing it, not knowing which try is right, is selection; selection is capped, and past a point extra samples only make the model surer of a confident mistake, even as every d…

Benchmarks are widely used to evaluate task completion by Large Language Models (LLMs), but this approach has accumulated construction-validity problems, and a passing score may not show whether the requested task was delivered. We study both problems. In a controlled code-as-spec setup, two production Copilot CLI agents (claude-opus-4.7, gpt-5.5) re-implement a React Fluent-UI data table in Angular as a reusable library under a hidden 222-test Playwright oracle across 18 runs and three oracle-availability conditions. Alongside the score, we run a mechanical library audit and check each ver…

As AI agents become increasingly capable of complex, long-horizon reasoning, rigorous and holistic evaluation is essential for measuring progress toward real-world healthcare applications. We introduce HealthAgentBench, a suite of 54 agentic healthcare tasks across 7 categories each with its unique environment. The benchmark suite spans diverse workflows throughout the patient journey and a broad range of modalities. Each task is designed to replicate an end-to-end clinical workflow: given minimal instructions, an agent must explore raw healthcare data, operate within a complex environment,…

Repository-level performance-optimization benchmarks such as GSO, SWE-Perf and SWE-fficiency evaluate coding agents by applying patches to real repositories and comparing runtime against unoptimized baselines and official reference patches. Their leaderboard scores are increasingly used as evidence of coding-agent progress, but those scores can conflate runtime instability, benchmark-specific scoring rules, and how many tasks are already solved by at least one public submission. We audit these issues across the three benchmarks. First, we replay the official reference patches for 740 code o…