All News

1525 technical news articles

Compile by Training Compiles Natural-Language Specifications into Reusable Offline Neural Functions

A new method first uses a teacher model to synthesize task data, then fine-tunes adapters for a small interpreter, packaging text specifications as versionable `.paw` functions. It achieves 83.6% semantic accuracy on difficult, ambiguous text tasks, but compilation requires an external model and roughly 40GB of accelerator memory, and it cannot provide the formal guarantees of conventional programs.

2026-09-07

代理工具

Prime Agent 0.9.2 Persists Trace Upload Progress with an On-Disk Cursor and Resumes Unfinished Work After Restart

The new release turns agent trace uploads into a persistent outbox. Processes no longer wait for network transfers when exiting, and can resume uploads from a byte cursor after restarting. It also begins recording semantic edges across sub-agents, but there is currently no corresponding delivery endpoint, so this should not be considered a complete distributed tracing capability.

2026-09-07

AI 安全

GitHub Agentic Workflows 0.88.4 Extends Data-Flow Policies to GitHub App Authentication

The new release automatically generates data-flow integrity and confidentiality policies for agentic workflows using GitHub App identities and adds sensitivity controls for trusted enclaves. Some isolation architecture changes exist only in recently merged main-branch code and should not all be treated as features delivered in 0.88.4.

2026-09-07

Soup 0.74 Fixes the fp32 Loading Path for Frozen Models, Cutting Peak VRAM to 18.7 GiB in a Single LoRA Test

Soup’s text, vision, and audio loaders did not specify a dtype, causing base models excluded from optimization to remain resident in VRAM as fp32. After the fix, the project’s H100 test fell from 48,241 MiB to 18,658 MiB, but the 2.59× improvement cannot yet be generalized to all models and hardware.

2026-09-07

AI 安全

HookPry Hijacks Updates to Compromise Seven AI Agents, Turning Lifecycle Hooks into an Execution Channel Outside Model Guardrails

Researchers were able to make previously trusted agent plugins execute malicious commands with host privileges by modifying only plugin metadata and hook configurations. All seven agents covered by 1,000 experiments were compromised at least once, but the success-rate and antivirus false-negative findings have not been independently reproduced.

2026-09-07

應用研究

WeatherNext 3 Directly Ingests Satellite Observations, Relaunching Global Probabilistic Weather Forecasts Every Hour

Google DeepMind’s new model no longer merely imitates reanalysis data produced every six hours; it incorporates low-latency satellite imagery into forecast initial conditions. It improves precipitation and station-level predictions, but evaluations in mountainous and polar regions and for heavy rainfall still expose limitations caused by sparse data and imperfect metrics.

2026-09-07

GitHub Repo

DLSS 5 Swapper 2.2.1 Brings Neural Rendering to Older Games via an Injection Layer, but Compatibility and Supply-Chain Risks Remain High

The new open-source tool uses ReShade, RenoDX, and Feeder to automatically deploy neural-rendering components in games without native DLSS support. The latest release adds an in-game overlay and per-game API fixes, but unsigned executables, modified DLLs, and anti-cheat systems mean it remains an experimental tool.

2026-09-07

AI 安全

OpenClaw 2026.9.2 Enables Cross-Agent Session Access by Default, Potentially Expanding Conversation Visibility After Upgrades

The new release gives eligible agents Swarm collaboration capabilities by default and can resume some interrupted tasks after a restart. More importantly, two previously unspecified settings now permit cross-agent reading and communication, requiring operators of shared Gateway deployments to review their permissions.

2026-09-07

GitHub Repo

OKF Agent Memory Stores Agent Memory in Git, but Independent Testing Finds Search Latency and Ranking Stability Issues

The new open-source tool uses Markdown, YAML, MCP, and local text retrieval to make coding agents’ long-term memory reviewable through Git. Its lightweight architecture works, but independent testing found that the “under 300 microseconds” claim applies only to extremely small datasets and that the same query may return results in different orders.

2026-09-07

開發者平台

OpenAI Java SDK 4.58 Recognizes Incomplete Web Search Calls, Rounding Out Responses Status Handling

OpenAI’s official Java SDK 4.58.0 adds support for recognizing the `incomplete` status of web search calls, preventing applications from treating interrupted search steps as successful or unknown values. The release is available on Maven Central, but projects still using the discontinued Spring Boot 2 starter cannot directly adopt the updated package.

2026-09-07

代理工具

RoutineKit 0.2 Packages Successful MCP Tool Sequences into Reviewable, Replayable Tools

RoutineKit, an independent open-source project, can turn a successful sequence of agent tool calls into a named tool with typed inputs, then replay it with new parameters. It preserves human approval and host safeguards, but currently supports only linear workflows and local stdio MCP—not general-purpose agent recording.

2026-09-07

推論系統

Speculative Macro Commit Pre-Executes Agent Toolchains, Cutting AppWorld Wall-Clock Time by 44.9%

New research enables a small draft model to pre-run multi-step tool operations in isolated environment snapshots, then commit the entire sequence once a large authoritative model confirms the first step. The method reduces latency on two agent benchmarks, but requires three GPUs and lowers AppWorld task completion from 41.67% to 40.48%.

2026-09-06