Builds structured knowledge graphs for retrieval-augmented generation via a multi-step GraphRAG pipeline that separates extraction from consolidation. Key features include typed two-stage extraction, DBSCAN-backed deduplication, LLM summarization, Leiden community detection, and a compact 7B extractor model (Meno-Lite-0.1).
AI Dataset2026
Provides behavior-preserving next-step training traces from Claude Fable 5 for supervised fine-tuning and analysis of instruction-following, tool-calling, and coding agents. Runtime-normalized, independently verified, and supplied as Parquet/JSONL with 13,357 cumulative rows from 2,443 accepted trajectories.
AI Dataset2026
Provides live Codex-CLI agent run traces from GPT-5.6 Sol capturing coding, debugging, security reviews, and harness/seed workflows in cumulative next-action prefixes — suitable for supervised fine-tuning and analysis of tool-using coding agents.
AI Dataset2026
Behavior-preserving dataset of GLM 5.2 coding and debugging agent trajectories for supervised fine-tuning and analysis; contains 1,821 cumulative next-step rows from 207 verified trajectories with multi-turn tool use, build-test-fix loops, and runtime-normalized traces.
AI Agent Papers2026
Converts completed on-policy trajectories into natural-language 'hindsight skills' and converts the skill-induced action probability shifts into a dense token-level on-policy distillation signal, jointly optimized with outcome-based RL to improve sample efficiency and long-horizon agent behavior.
Embodied AI2026
Analyzes adversarial weaknesses of World-Action Models (WAMs) via BadWAM, a framework that crafts visual perturbations to decouple a model’s imagined future from its executed actions. Introduces two attack modes—action-only (disruptive) and imagination-preserving (stealthy)—and shows large drops in closed-loop task success (e.g., 96.5%→43.1%).
AI Agent Papers2026
A vision-language-action foundation model trained on 100k+ hours of real-world robot manipulation trajectories to follow natural-language instructions and adapt to downstream tasks with minimal fine-tuning. Uses a two-stage (pre-/post-) training recipe and a scalable auto-labeling pipeline; shows clear scaling benefits and state-of-the-art sim-to-real transfer on standard benchmarks.
AI Agent Papers2026
Guides an LLM agent to build persistent, editable DAG-based data pipelines via typed, incremental mutations instead of free-form scripts. Combines DataFlow-Skills, a Model Context Protocol exposing live operator registry and pipeline state, and a synchronized Web UI; achieves 93.3% end-to-end pass rate on a 12-task benchmark while cutting cost and latency versus script baselines.
AI Agent Papers2026
Prunes tool-output lines inside a coding LLM agent by turning the agent's own internal representations into per-line keep-or-prune labels. Implements a small classification head plus a length-aware embedding, saving up to 39% of tokens across benchmarks while preserving task quality.
AI Model2026
Compact 3B-scale agentic LLM for multi-step tool use and reasoning, using a Looped Transformer to increase capacity without adding parameters; built for local deployment with configurable "thinking" modes and benchmark gains vs larger open models.
A PyTorch-native training framework for agentic reinforcement learning research that keeps researcher-facing code compact and editable. Uses an asynchronous loop to train multimodal and mixture-of-experts policies while never training on tokens the agent didn't generate; matches Megatron-style stacks under a comparable protocol and ships recipes and containers on GitHub.
AI Dataset2026
A curated collection of 49,772 teacher-generated chat traces from qwen3.8-max-preview for supervised fine-tuning and off-policy distillation. Preserves visible chain-of-thought blocks, emphasizes math/code/reasoning mixes, and includes provenance and licensing cautions tied to Alibaba Cloud Model Studio.