GitHub Repo

GitHub 熱門專案、開源工具、框架、runtime 與重要 release。

Optuna 5.0 Changes Its Default Search Strategy, While Rustuna Tackles Framework Overhead in Large-Scale Trials

In its first major change to the default sampler since version 1.5, Optuna 5.0 adopts multivariate TPE and Constant Liar for single-objective search, while replacing NSGA-II with TPE for multi-objective search. Released alongside it, Rustuna reimplements a compatible core in Rust to target sampling and storage bottlenecks in workloads involving hundreds of thousands of fast evaluations.

2026-09-08

代理工具

Ripwire 0.4 Uses a Code Graph to Narrow the Search Scope for Agents, but Saving Tokens May Not Reduce Task Costs

Ripwire uses tree-sitter, a symbol relationship graph, and Personalized PageRank to generate reproducible file rankings and change-impact scopes for coding agents. First-party tests show better localization rates than five comparison tools, but an independent review found that compressing context may still degrade answer quality and increase overall agent costs.

2026-09-08

代理工具

Hermes Agent 0.21.1 Sweeps Up More Than 5,000 Commits in One Week, Making the Patch Release Closer to an Upgrade Requiring Full Revalidation

Nous Research has delivered a massive week’s worth of main-branch changes under the 0.21.1 patch label, covering MCP authorization, scheduling, desktop sessions, and subagent reliability. The company has not provided an itemized changelog or performance data, so production deployments should not assume limited compatibility risks simply because it is labeled a “patch release.”

2026-09-08

最新模型

EXAONE Finance Releases 202M-Parameter Attention-Free Time-Series Model, with Separate Licenses for Weights and Code

LG AI Research replaces self-attention with causal convolutions and a group-aware MLP, allowing computational costs to scale linearly with sequence length and the number of variables. The model outperformed 43 baselines on the team’s own FinVerse benchmark, but noncommercial model weights and independently unreproduced investment performance limit its immediate deployment value.

2026-09-08

代理工具

Qwen Code Preview Advances Multi-Agent Workflows with an Observable, Isolated Environment for Long-Running Tasks

Qwen Code 0.23.1-preview.2 adds workflow visualization, real-time sub-agent status, worktree-isolated tasks, and a session resource directory. The new release also uses leases and message-rejection mechanisms to handle concurrent writes, but it remains a preview lacking performance and reliability data.

2026-09-08

推論系統

KVMem Pages Agent History to Memory and NVMe, Letting a 24GB Laptop GPU Address a Million-Token Workspace

Instead of compressing overflow agent history into summaries, KVMem preserves precomputed KV states and loads relevant blocks for each query. The authors demonstrate a logical million-token workspace using Qwen 27B on an RTX 5090 Laptop GPU, though model support and performance figures remain closely tied to their custom runtime.

2026-09-08

推論系統

ACE Skips Low-Contribution MoE Experts Per Token, Accelerating Prefill by Up to 2.25× on A100

ACE estimates expert contributions from SwiGLU weights and router directions without retraining the model or using calibration data. The authors achieved better quality at high skip rates across three MoE models, though the performance results remain limited to a single A100 and thresholds matched to the workload.

2026-09-08

AI 基礎設施

Megatron-LM Expands Prefix-Aware Routing and Dynamic Batching, Enabling Hybrid Model Inference to Share KV Cache and Mamba States

NVIDIA summarized Megatron-LM changes from the first week of September, advancing the data-parallel inference frontend, prefix caching, and variable-length sequence processing toward a more complete serving path. The new features improve state reuse for hybrid Transformer/SSM models, but the company has not yet published throughput, time-to-first-token, or cross-node scaling data.

2026-09-08

代理工具

τ^τ-Bench Puts “Building Agents” Itself Under Evaluation; Best Coding Agent Reaches Just 30% of Expert Reference Performance

Sierra’s new benchmark no longer tests only finished agents. Instead, it requires coding agents to build deployable customer-service systems from fragmented business materials, customer interviews, and existing code. The best combination passes just 23.9% of hidden simulated traffic, exposing shortcomings in requirements discovery, architecture exploration, and validation.

2026-09-08

最新模型

Tencent’s EVIE Cuts Visual Document Retrieval Index Costs with Truncatable Multi-Vectors and Token Clustering

Tencent has released EVIE-8B and EVIE-4.5B, which create per-token representations directly from page images instead of performing OCR before retrieval. The 4.5B model supports runtime-adjustable vector dimensions and can compress each page to 32 vectors, although its leading results currently come mainly from the team’s own evaluations.

2026-09-08

AI 安全

OpenMAIC 1.0.1 Patches Four Vulnerabilities; Unauthenticated Model Endpoints Could Be Used to Access Cloud Metadata

Tsinghua University’s multi-agent classroom platform OpenMAIC has patched an SSRF vulnerability that could be triggered without authentication, along with stored XSS, path traversal, and redirect-validation flaws. The upgrade also tightens settings for local models and development-time validation, meaning self-hosted users cannot simply replace the container while ignoring environment variables.

2026-09-08

最新模型

MiniCPM5-2B Opens Its Weights and Training Data, Challenging On-Device Agent Workloads With 2.5B Parameters

OpenBMB has released MiniCPM5-2B with a 131K context window, alongside approximately 86,000 verifiable reinforcement learning examples. Its official aggregate score exceeds those of some 4B models in the comparison table, but the evaluations were conducted by the model’s developer and await independent reproduction.

2026-09-08