Archive
新模型、開放權重、多模態、語音、影像與模型架構。
推論系統
Architecture support merged on September 14 allows Maple-Preview to be loaded and converted without relying on DeepGrove’s llama.cpp fork. The model activates only about 1 billion parameters per token, but GPU backends, quality evaluations, and official speed claims still require separate verification.
代理工具
New experiments show that relative terms such as `previous` in “draft–validate–revise” pipelines may change referents as model roles shift. Increasing reasoning effort can dramatically improve some models, but model selection matters more than simply adding test-time compute.
最新模型
SteerDuplex brings tone, accent, speaking rate, and interruption handling into a unified post-training framework for full-duplex speech, while introducing SteerBench, which contains 390 spoken prompts. The paper reports substantial gains but also reveals that the model can game interruption metrics by ending its responses prematurely.
GitHub Repo
OpenArch rewrites models including Llama, DeepSeek, Gemma, Qwen, and Kimi as architecture-specific, pure PyTorch examples, making it easier to compare attention mechanisms, normalization methods, and MoE routing. The project quickly gained attention on Hacker News on September 14, but most models remain unfinished, and the existing code explicitly does not target performance or compatibility.
AI 安全
A community audit found that `huggingface_hub` identifies agents such as Codex, Claude Code, and Cursor through environment variables and adds their names to the User-Agent header of Hub requests. The mechanism has long been publicly documented by Hugging Face, but the documentation still does not clearly explain whether general telemetry opt-out settings fully cover agent labeling.
推論系統
The new open-source benchmark runs small models on Jetson devices, Macs, Raspberry Pis, and Android devices, measuring decoding speed, tokens per joule, latency, and temperature instead of focusing solely on data-center GPU throughput. Its creator says roughly 1,000 Jetson configurations have been tested, but the public leaderboard currently loads no data, leaving users to verify results through repository reports and raw artifacts.
最新模型
Shanghai AI Laboratory has upgraded Intern-S2-Preview-397B to a production release, adding long-horizon agent capabilities and releasing BF16 and FP8 weights under Apache 2.0. API users must migrate to `intern-s2`, but the official documentation contains conflicting information about the latest alias, so users should verify the actual model version returned during migration.
GitHub Repo
The September 13–14 llama.cpp prerelease builds add a CUDA F32 path for systems where BF16 is unavailable, preventing outright failures on some older GPUs. Another fix reports an error when Nemotron-H NextN/MTP metadata is incomplete instead of triggering SIGFPE while loading the model.
推論系統
Colibrì, which has recently gained traction in the community, uses a pure C engine to load sparse experts on demand, enabling models with 744B to 2.8T parameters to run on non-data-center hardware. What it lowers is the resident-memory requirement; actual decoding speed remains heavily constrained by SSD bandwidth, expert hit rates, and quantization quality.
AI evaluation / scientific reasoning
A human audit covering six physics benchmarks found that most answers initially marked wrong actually involved incorrect reference solutions, grader flaws, or incomplete problem statements. After problematic items were corrected or removed, GPT‑5.6‑Sol achieved 94.4% pass@4 on the retained CritPt problems—but that is not equivalent to its score on the original full leaderboard.
最新模型
OpenBMB has released MiniCPM5-2B, its training data, and multi-stage weights spanning pretraining through RL, along with GGUF, MLX, GPTQ, and speculative decoding variants. The model retains the standard Llama architecture and a 128K context window, but the SGLang parser required for tool calling may still need to be installed from the main branch.
代理工具
The experimental main branch of Astah Pro MCP adds programmatic tool calling, replacing nearly 400 fine-grained operations in direct mode with 83 tools to reduce the burden of loading large tool catalogs for agents. The changes do not yet appear in Astah’s official stable release information for version 0.2.9, and model data, logs, and editing permissions are still exposed to external agents.