All News

1537 technical news articles

Global Debugging Memory Helps AutoResearch Agents Avoid Repeating Mistakes, Raising AIDE Gold-Medal Runs from 22 to 38

With GPT‑5‑mini fixed as the model and a two-hour budget, researchers enabled a debugging advisor to share execution-environment constraints across search branches, allowing all 90 AIDE runs to produce valid submissions. The results suggest that agent scaffolds still contain substantial recoverable compute, though the experiments covered only nine tabular prediction competitions.

2026-08-12

代理工具

gpt-oss-20b Finds 13 Classes of Defects in Rope Refactoring Diffs, 12 Accepted by Maintainers

Researchers had a local model inspect only before-and-after Python refactoring diffs and identify 13 classes of behavioral changes across 217 Rope transformations. Maintainers accepted 12 as defects, but the model’s precision was only 0.67—too low to replace testing or static analysis.

2026-08-12

XCoT-VLA Compresses Autonomous-Driving Reasoning into 2–6 Action Tokens, with Interface Latency as Low as 38.6 ms on H100

XPeng introduces XCoT-VLA, replacing dozens of natural-language reasoning tokens with a limited set of semantic action tokens, then using flow matching to generate continuous trajectories. Open-loop tests show lower lateral error, but no closed-loop safety evaluation has been reported, and the latency figures do not cover the full in-vehicle hardware and software stack.

2026-08-12

AI 安全

REDAgentBench Recalculates Agent Attack Rates Using Final Service States, Finding Text Trajectories Miss Up to 11.72 Percentage Points

REDAgentBench executed 1,661 attack cases across sandboxed workspace, email, browser, banking, and external file services, rather than judging harm solely from an agent’s response. When the same runs were scored using service receipts and final states, the attack success rates of all six models were 7.73 to 11.72 percentage points higher than with trajectory-only scoring.

2026-08-12

代理工具

Agent instruction files grow by an average of 226%; study curbs CLAUDE.md bloat with “rule rationales”

A study of 1,867 public repositories found that agent instruction files such as CLAUDE.md continually accumulate rules while older rules are deleted less and less often. Controlled experiments showed that retaining model-readable rationales for adding rules can eliminate 99.3% of superfluous instructions, though the results have yet to be prospectively validated in large real-world codebases.

2026-08-12

AI 軟體與企業管理

Microsoft Adds Separate Removal and Sign-In Controls After OneDrive Photos Was Accidentally Pushed to Enterprise PCs

The OneDrive photo experience, still in internal testing, was accidentally enabled more broadly, exposing an entry point on some managed Windows 11 PCs. Initially, it could not be removed separately while retaining the sync client. Microsoft has now added administrator sign-in policies and standalone uninstallation to the Microsoft 365 Roadmap, but both fixes remain scheduled for future release.

2026-08-12

MeViS-Audio Champion Replaces Single-Model Decisions with Candidate Mask Consensus

The winning report for MeViSv2-Audio at the eighth LSVOS breaks speech transcription, visual grounding, segmentation, and target-presence detection into a multistage pipeline, then selects a trajectory based on agreement among candidate masks. The system achieved a final score of 0.769589, although the official public leaderboard still says results are to be announced, and neither the code nor complete ablation data has been released.

2026-08-12

代理工具

OEO Lets GPT‑5.5 Orchestrate Its Own Agent Improvement Process, Winning 12 of 14 Comparisons

Open-Ended Optimization fixes the objective, data boundaries, interaction permissions, and evaluation, but does not prescribe how the agent should gather evidence or modify skills. With GPT‑5.5 as the optimizer, it outperformed SkillOpt- and GEPA-style workflows in most settings, but medium-strength and weaker models could not maintain the same advantage.

2026-08-12

AI 安全

Only 10 of 30 LLM Vulnerability Validation Artifacts Passed Patched-Version Falsification After Reruns

A preregistered reproducibility audit found that a vulnerability PoC that executes and triggers a crash does not necessarily reproduce the specified CVE. In 20 of the 30 evaluable cases, the same signal still appeared in the patched version, showing that agentic security benchmarks need stricter oracles than strings and exit codes.

2026-08-12

代理工具

DSLE Packages 22 Dark Souls Boss Fights as Gymnasium Environments, but the Public Repository Is Still an Empty Shell

Using Docker, Wine, virtual displays, and process-memory reads, DSLE turns real-time combat in a commercial game into a reinforcement learning benchmark with parallel resets. Five baselines consistently conquer only the tutorial boss, while the code the paper claims is public has yet to appear in the repository.

2026-08-11

應用研究／AI 基礎設施

GENCO Uses a Single Graph Neural Network for Three Power-Grid Solvers, Boosting Large-Scale AC Power-Flow Throughput by About 29×

Through heterogeneous graph attention, physical-variable decoding, and iterative residual correction, GENCO unifies power flow, optimal power flow, and state estimation. Batched inference on large grids is substantially faster than conventional AC solvers, but its zero-shot cross-grid and real-world SCADA accuracy remains insufficient to directly replace existing tools.

2026-08-11

AI 編程工具

Coderlet Open-Sources a Readable Control Layer for Coding Agents, but Lacks Sandboxing and Automated Tests

Coderlet separates model calls, tool validation and execution results, and persistent memory into three explicit boundaries, providing an executable Python implementation under the MIT License. It is suited to studying agent control flow and failure recovery—not a higher-performing coding agent—and currently operates directly with the user’s operating-system permissions.

2026-08-11