Proposes Monotonic Inference Policy Improvement (MIPI) and a two-step Monotonic Inference Policy Update (MIPU) to address training–inference probability mismatch in LLM reinforcement learning by constructing sampler-referenced candidate updates and accepting synchronized updates using an inference-gap proxy; shows improved reasoning accuracy and stability under FP8-quantized rollouts.
AI Dataset2026
Provides synchronized four-perspective Rocket League match recordings with per-frame H.264 video, player action streams, event logs, and privileged physics state — released as WebDataset shards in a ~4,000-hour slice (1,000 match-hours × 4 perspectives). Includes 720p@20fps video, multi-hot keyboard actions, and CC BY-NC-SA-4.0 license.
AI Dataset2026
Provides ~494.7 hours of trimmed native PC/console gameplay screen recordings organized by game, with per-session clips plus input and per-frame event annotations. Each workflow includes clip.mp4, events.json, frame_events.json, and metadata — suitable for training vision-action, behavior-cloning, and gameplay understanding models.
Trains a transformer-based graph encoder with RL-guided adaptive masking so retrieved subgraphs embed relationships that better align with frozen LLM text encoders, improving GraphRAG performance with non-parametric retrievers on GraphQA benchmarks.
Zhilin Wang, Han Song +14University of Science and Technology of China, The Chinese University of Hong Kong +6
Provides a benchmark and protocol to evaluate agents that iteratively edit executable policies under a fixed interaction budget, recording full execution–feedback–revise trajectories. Built from compact RL environments with trajectory-level diagnostics and hidden held-out validation.
Embodied AI2026
Provides a systematic benchmark and design roadmap for video-based world models to evaluate robot policies, introducing WMBench and GigaWorld-1 optimized for long-horizon, action-faithful rollouts. Offers controlled comparisons across model families, action encodings, and 324k+ simulated vs real rollouts, with code, models, and datasets released for reproducible evaluation.
AI Agent Papers2026
Introduces a bounded-memory, typed-retrieval contract for long-horizon LLM agents and evaluates it in Slay the Spire 2 — assembling per-decision prompts from five typed slots rather than appending raw transcripts. Key outputs include ablationable memory layers, 298 labeled trajectories, and reproducible analysis scripts.
Embodied AI2026
Detects when an action-chunked VLA policy drifts from expected visual dynamics and triggers lightweight corrective replanning via a latent-space vision monitor and online gradient guidance; creates an event-driven adaptive action horizon without retraining the backbone.
Embodied AI2026
Evaluates generalist robot manipulation policies across simulation and real-world settings using 42 sim tasks and 18 real tasks; measures generalization, memory, precision, long-horizon execution and open-vocabulary instruction following, and provides a cloud-accessible real-world evaluation system with XPolicyLab integration and a public leaderboard.
AI Agent Papers2026
Trains cross-platform GUI agents by combining a Uni-GUI cross-platform dataset with platform-conditioned multi-teacher on-policy distillation, enabling a shared policy to adapt to new platforms while retaining platform-specific behaviors; suitable for research on continual GUI agent learning and cross-platform adaptation.
Stabilizes on-policy policy distillation by dynamically constructing a proximal teacher that controls gradient variance. Provides theoretical global convergence and monotonic improvement bounds, and shows improved training stability, sample efficiency, and final performance on mathematical reasoning tasks with zero extra compute overhead.
Transfers RL-induced policy shifts from a smaller 'weak' teacher to a stronger target by using the teacher's post-/pre-RL log-ratio as a dense implicit reward applied on the student's on-policy states. Enables reuse of RL supervision without running RL rollouts on the target, improving sample/time efficiency.