Archive
新模型、開放權重、多模態、語音、影像與模型架構。
最新模型
XPeng uses behavioral probes, cross-scale distillation, and LoRA to shrink the Qwen3-ASR-0.6B audio encoder from 18 layers to 14. The weights and inference code are public, but the full distillation pipeline and more than 280,000 hours of training data have not been released.
推論系統
The new release serves MiniCPM-o 4.5, PersonaPlex, and Nemotron VoiceChat through a shared WebSocket runtime, while adding diffusion batching, parallelism, and component offloading. It remains a pre-release, and some endpoints silently fall back to turn-based voice, so a successful connection alone does not confirm full-duplex capability.
AI research
The new method arranges each recurrent-network update into a continuous flow from noise toward an answer, allowing the model to learn state across steps without fully backpropagating through every iteration. At inference time, more integration steps can be used for additional compute, while different initial noise samples can produce multiple candidate solutions.
最新模型
Sakana AI has launched Fugu Max and Fugu Ultra v2, optimizing multi-model orchestration along two separate tracks: lower cost and greater capability. Both retain an OpenAI-compatible interface, but the official benchmarks still combine Sakana’s own tests with vendor-reported results, providing insufficient evidence of cost advantages in real-world workloads.
AI 安全
Anthropic’s new threat report shows that several types of attackers have upgraded Claude from a coding assistant into a schedulable attack-orchestration layer with persistent memory. Some workflows automatically modify, rebuild, and redeploy malware after detection, although humans still largely control target selection and the review of stolen data.
最新模型
OpenAI is giving developers access to GPT‑Live‑1, which can listen, speak, and handle interruptions simultaneously, while allowing the voice front end to delegate complex reasoning and tool operations to other models. The new architecture reduces fragile handoffs between STT, LLM, and TTS, but backend reasoning, tools, and call infrastructure are billed separately.
AI for Science
ARCHE integrates literature retrieval, competing hypotheses, electronic structure calculations, and reflective review into a resumable closed loop, while releasing 65 actions and data from three case studies. However, the archived data already differ from figures reported in the paper, and the specialized model weights, training data, and some chemical evidence were not released with the code.
代理工具
Auto-RecSys links research ideas, distributed training, failure recovery, and results analysis in a dual-loop system that persists across sessions. In 31 internal experiments, operational fixes fell from 4.0 to 0.5 per run, but Meta has not yet shown that the agent’s proposed model changes can consistently improve recommendation quality.
最新模型
NVIDIA has released the weights for two Nemotron 3 Ultra math specialists, their training data, and a resumable natural-language proof pipeline. The system’s official score was 30 out of 42. The results also expose a shared blind spot among model-based verifiers: internal evaluations overestimated performance on two problems, while high precision came at the cost of extensive false rejections and substantial compute.
GitHub Repo
The new open-source framework uses predictive routing to preload experts for the next step and Recover-LoRA to compensate for int4 quantization loss, so the entire 23GB checkpoint does not need to remain resident in memory. Official M4 Pro tests measured 14.9–17.7 tokens per second, but the framework currently supports only Apple Silicon, and its memory and quality results have yet to be independently reproduced.
最新模型
Microsoft has updated its managed LLM Speech offering, emphasizing recognition of mixed-language speech, proper names, and numbers, while claiming up to a threefold latency reduction versus version 2605 in some scenarios. The new model uses the existing Fast and real-time APIs but is deployed automatically on the server side, so teams operating existing voice agents must establish their own regression tests and version monitoring.
推論系統
Dynamo can now assign visual encoding, LLM prefill, and decoding to separate workers and different GPU tiers, preventing image requests from blocking text-only traffic. Although NVIDIA’s tests show improvements of up to 5× in time to first token and 7× in end-to-end latency, long outputs or large dense models may offset the benefits of disaggregation.