AI Coding Live News

Shanghai AI Laboratory has moved Atria Dawn from Preview to a full release, with open weights (plus an FP8 build) on Hugging Face under internlm/Atria-Dawn and on ModelScope, MIT licensed. Built on the 744B-parameter MoE GLM-5.2, it targets long-horizon research and engineering agent work; context doubles from the Preview's 256K to 512K and input is text only. The vendor table shows it leading the listed models on AutomationBench, CyberGym and SkillsBench, while still trailing Claude Opus 5 on coding tests such as SWE-bench Pro and Terminal-Bench 2.1 (all vendor-reported).

Zoom

The Shanghai Artificial Intelligence Laboratory published the technical report for Sage-A1.5 on arXiv:2610.28578, introducing an open-weights 35B Mixture-of-Experts (MoE) foundation model purpose-built for agentic tasks. Building on the Agents-A1 lineage, Sage-A1.5 focuses on scaling the agent execution horizon rather than model parameters through adaptive capability composition. Tailored for complex repository-level software engineering and multi-step scientific hypothesis exploration, Sage-A1.5 consistently outperforms dense 70B+ baselines on long-horizon reasoning benchmarks while maintaining lightweight deployment footprints.

Zoom

Advancing autonomous AI-for-AI (AI4AI) engineering, researchers from Shanghai AI Lab (InternScience) release MLExpert (arXiv:2610.22344), a 35B Mixture-of-Experts agentic model for autonomous machine learning engineering. Powered by the Kaggle-MLE Factory, hierarchical agentic SFT, and Windowed Interaction RL (WIRL), MLExpert boosts average medal rates on MLE-Bench Lite from 37.9% to 72.7%—matching frontier giants like Kimi-K3 and GLM-5.3 while generalizing to agent scaffold optimization.

Zoom

Shanghai AI Lab announced on 2026-10-15 that Disheng Xu (Great Bay University) and Jiesong Zhang (KTH), aided by an internal version of Intern-S2, proved the codimension-one case of Ghys's 1995 holomorphic Anosov conjecture, including complex dimension three (arXiv:2609.37695). Peking University's team formalised all four main theorems in Lean with the SHEAF framework in about five days with up to 100 concurrent agents and ~550k lines, sorry-free and accepted by Lean comparator, with Hsiang–Wall (1969) as an explicit hypothesis.

Zoom

Vision-language navigation requires abundant, high-quality language supervision, yet traditional synthetic pipelines generate supervision from static models that lag behind evolving navigation policies. Shanghai AI Lab, Tsinghua, and SJTU introduce RSI-Nav, an embodied recursive self-improvement framework combining instruction-following and instruction-generation in a single model. Using shard-wise dual bootstrapping on unlabeled trajectories, RSI-Nav achieves state-of-the-art 68.6% SR on R2R-CE and 72.4% SR on RxR-CE without human intervention.

Zoom

Empowering AI agents to autonomously achieve Recursive Self-Improvement (RSI) through ongoing environmental interaction is a cornerstone of artificial general intelligence. However, practical RSI pipelines are severely constrained by two fundamental bottlenecks: scarce and expensive ground-truth feedback, and the degradation of predictive world models as evolving policies induce out-of-distribution drift. Researchers from HKUST, Shanghai AI Laboratory, Peking University, and CUHK introduce RSI-Anything (arXiv:2610.13912, code: github.com/Gen-Verse/RSI-Anything), a domain-general framework unlocking recursive self-improvement via predictive world models. Built upon the JEPA-Anything architecture, RSI-Anything establishes a disciplined six-stage paradigm: Propose → Predict → Select → Evaluate → Learn → Retain. By enforcing rigorous compatibility and safety gates governing policy-world model retention, the system continuously converts interaction logs into persistent capability upgrades. Across diverse domains, RSI-Anything drives substantial performance gains: boosting automated code repair success from a 26.3% baseline to 31.2% (+4.9 points), reducing GUI navigation action footprints by 39.4%, more than doubling laboratory planning scores from 21.80% to 46.01% over five iterations, and achieving a 96.15% chemical yield, with codebase fully open-sourced.

Zoom

Streaming video LLMs must continuously capture temporal evidence before its downstream relevance is known and proactively respond when sufficient evidence arrives. Balancing real-time perception with reusable factual memory has remained a core challenge: revisiting raw visual features explodes memory, while lossy feature pooling destroys critical spatiotemporal context. Researchers from Nanjing University and Shanghai AI Lab introduce OneStreamer (arXiv:2610.01762, ~200 upvotes on Hugging Face). The framework introduces Proactive Hierarchical Caption Memory (PHCM) to generate time-grounded captions and event summaries on the fly, paired with Proactive State Transition Learning (PSTL) to prevent idle waiting frames from overwhelming optimization gradients. Built on the newly released OneStreamer-1M dataset, their 4B model establishes state-of-the-art results across all eight streaming video benchmarks.

Zoom

Shanghai AI Lab’s LMDeploy shipped v0.18.0 on 2026-09-28: input logprobs, expanded SM90 quantized GEMMs with unified TurboMind linear paths, Ascend GLM-5.2, Qwen3.5 dflash, and MoE shared-expert/FFN sharding. PyTorch path prototypes TurboMind W4A16 (AWQ), piecewise CUDA Graph prefill, XTuner TileLang sparse MLA, checkpoint-engine weight updates, and request-only KV cache metrics.

Zoom

Shanghai AI Lab and OpenDataLab open-sourced InternW0-Δ, a unified World Action Model (WAM) pretrained on over 20,000 hours of curated robotic demonstrations, UMI data, and egocentric videos. Built on a Mixture-of-Transformers (MoT) architecture that binds visual predictive dynamics with robot actions, it introduces 'Causal Imprint' to inject future representations directly into the action expert without requiring costly future-video rollouts during inference.

Zoom

Shanghai AI Laboratory has open-sourced AV-GRPO (arXiv: 2609.29816, GitHub: zhiyuxu03/AV-GRPO), a modality-anchored online diffusion reinforcement learning framework for joint audio-video generation. Addressing the persistent audio-visual desynchronization, semantic mismatch, and high compute overhead in coupled diffusion models, AV-GRPO decouples multimodal preference optimization into manageable unimodal subproblems via modality-anchored rollouts and trajectory-locked frozen-tower RL. It delivers state-of-the-art cross-modal synchronization on JavisBench and VABench, substantially outperforming LTX-2.3.

Zoom