AI Model2026
A 9B agentic multimodal SFT checkpoint distilled from Qwen3.5-9B for coding, general agent tasks, visual coding and cybersecurity. Provided by Xiaomi MiMo as a research seed (77.4B-token SFT mix) to bootstrap agentic RL and tool-use experiments.
AI Dataset2026
Provides 5.5K+ self-contained data-analysis RL tasks: each row bundles a real tabular dataset, a question, and a deterministically-gradable gold answer. Verified from jupyter-agent notebooks; splits for training, held-out testing, and quick eval; intended for prompting, fine-tuning, and agent RL.
Adapts LLM agents online by self-distilling verified execution trajectories into persistent LoRA weights during deployment to improve success and efficiency on long‑horizon tasks. Uses a frozen stable copy as a privileged teacher to predict hindsight next‑token distributions and filters invalid-action turns so experience consolidates without external solutions or memory retrieval.
Provides a large corpus of runnable 3D programs and a multi-agent harness that generates, executes, measures, visually judges, and repairs executable 3D code to produce editable, verifiable 3D scenes. Includes ~269K runnable programs, 2.03M render–caption pairs, and an automated execution+repair pipeline.
AI Model2026
Coordinates long-horizon scientific and engineering tasks by delegating independent investigations to parallel subagents and revising its approach through evidence checks. It combines tool use and multimodal input in a 35B mixture-of-experts model.
AI Model2026
Uses reinforcement learning to improve task generation, solving strategies, and solution rollouts in a continuous loop. This 122B MoE model targets coding, reasoning, and agent tasks, supports a 262K context window, and can run on one 80GB GPU.
AI Agent2023
Memory layer that lets AI agents remember users and context across sessions.