最新模型

新模型、開放權重、多模態、語音、影像與模型架構。

Gemini 3.8 Flash Strengthens Long-Horizon Coding Agents, but Migration Requires Sampling and Thinking Parameter Changes

Google has moved Gemini 3.8 Flash into production, offering a 1 million-token context window, three thinking levels, and built-in tools, while making it the default model for Antigravity managed agents. Coding benchmark results have improved, but most figures still come from Google, and the model card also reveals a relative decline in non-English safety evaluations.

2026-09-05

AI 安全

Cloudflare Correlates Code Vulnerabilities With Live Traffic to Generate Patches and WAF Rules Through a Multi-Agent Workflow

Cloudflare’s early-preview service combines source code analysis results with actual routes, traffic, WAF status, and attack events, then uses GPT-5.6 Cyber to validate vulnerabilities and proposed fixes. The model does not directly modify the environment, but data is sent to OpenAI for inference, and Cloudflare has not disclosed false-positive or remediation success rates.

2026-09-05

電腦視覺

Scal3R Fixes Drift in Long-Video 3D Reconstruction With Multi-Reference Pose Queries, Training Only About 1% Additional Parameters

Scal3R retains a frozen CUT3R or STream3R backbone, using lightweight query tokens to estimate relative poses against multiple historical keyframes before passing them to online pose graph optimization. The authors report a mean ATE on KITTI more than 60% lower than the strongest online competitor, but the results use Sim(3) alignment and therefore should not be interpreted as preserving true scale.

2026-09-05

最新模型

VibeVoice-ASR-Streaming Combines Incremental Recognition and Speaker Attribution, Open-Sources 1.5B and 7B Weights

Microsoft Research’s streaming speech model interleaves audio chunks, brief lookahead, and the existing transcript to directly output “who said what.” The public release supports ten languages and custom hotwords, but expected speaker-attribution latency remains two seconds, and performance degrades during prolonged overlapping speech.

2026-09-05

推論系統

Declarative Attention Lets Models Specify KV Cache Read Ranges, Cutting Tokens Attended During Long-Context Decoding by Over 30%

KAIST and Google DeepMind have introduced Declarative Attention, which allows existing models to switch among global, focused, and local attention modes within their reasoning text. The authors reduced attended tokens by 31.1% to 52.0% across two models, though the speed gains remain model-based estimates and accuracy declined slightly.

2026-09-05

最新模型

GLM-5.3-Flash Uses Hybrid Sparse/Linear Attention to Cut Million-Token Inference Costs

Zhipu AI has unveiled GLM-5.3-Flash, a native multimodal model with 320B total parameters and 18B active parameters per token, releasing its weights under the MIT License. The new architecture is claimed to require three times less attention computation and 4.4 times less KV cache than GLM-5.3, but “Flash” does not mean it can be easily deployed on a typical workstation.

2026-09-04

推論系統

Uno Uses Discrete Diffusion to Assist Autoregressive Models, Enabling Parallel Decoding Without a Separate Draft Model

Uno attaches lightweight diffusion weights to an autoregressive LLM, then uses Ψ-Spec to propose multiple tokens at once while preserving the original model’s sampling distribution. The authors report speedups of up to 3×, though the current figures are limited to the research team’s specific models, hardware, and inference engine.

2026-09-04

GitHub Repo

Hugging Face Tokenizers 0.23.2 Brings Cross-Lingual Fairness Objectives to BPE Training

The new release adds Parity-aware BPE, enabling tokenizers to prioritize improving the language with the worst compression at each merge instead of merely pursuing the highest overall frequency. The feature can be used to train new vocabularies, but it does not automatically reduce Chinese token costs in existing models.

2026-09-04

最新模型

Runway GWM Worlds 2 Continuously Generates Interactive 720p Video with a Sliding KV Cache

GWM Worlds 2 combines autoregressive diffusion, a causal video decoder, and the structured WorldPrompt format to enable real-time control of persistent worlds through text-based actions and camera signals. The system remains a closed research preview, with fast camera movement, long-term consistency, and reproducible evaluation among its clear shortcomings.

2026-09-04

代理工具

DisCo Distills 1,000 Machine Learning Repositories Into 5,353 Verifiable Agent Skills

New research turns source code, documentation, and tests into operational knowledge that can be loaded on demand, delivering gains across four research benchmarks with the same underlying model and agent framework. Skill-building costs are accounted for separately, however, and the results remain the authors’ own evaluation, so they should not be treated as evidence that skills can replace model training at low cost.

2026-09-04

代理工具

Anthropic’s `ant apply` Brings Agents, Skills, and Memory Stores Under Declarative Version Control

The Claude Platform CLI adds `ant apply`, which synchronizes managed agent resources from Markdown, YAML, or JSON files in a repository. It provides previews, dependency ordering, and drift detection, but its state file still contains organization identifiers, and it currently manages only the Anthropic platform.

2026-09-04

代理工具

Cohere ATE Maps Nearly 700,000 MCP Tools to Occupational Tasks, Finds Only 2.6% Can Complete Them End to End

Cohere Labs has released the Agentic Task Ecosystem, which aggregates 696,291 tools from seven directories and maps them to tasks in the U.S. O*NET database. The results suggest that many MCP tools cover only fragments of a workflow, but the 2.6% figure comes from semantic matching and model-based classification—not real-world execution tests.

2026-09-04