…High Accuracy Agentic Post-Training at Low Compute Cost (2026) Prune as You Generate: Online Rollout Pruning for Faster and Better RLVR (2026) Learning Adaptive LLM Decoding (2026) Prompt replay: speeding up…
…Tolerable for human readers, these costs become critical when AI agents must understand, reproduce, and extend published work. We introduce the Agent-Native Research Artifact (ARA), a protocol that replaces the narrative…
…spanning computation, memory, communication, and cost across network tiers. We present collaborative inference along two complementary and composable dimensions: vertical device-cloud collaboration and horizontal multi-agent collaboration , which can be combined…
…This is a big step toward more capable AI agents. At Automatio.ai , we're working on integrating similar autonomous web agents to streamline data extraction and web automation, letting users build…
…We introduce a low-cost, automated, and scalable data acquisition pipeline for creating customized occupancy datasets , and demonstrate its effectiveness through MinecraftOcc, a large-scale dataset featuring 100,165 images from 156…
…Step-wise On-policy Distillation for Small Language Model Agents (2026) MAD-OPD: Breaking the Ceiling in On-Policy Distillation via Multi-Agent Debate (2026) OPSDL: On-Policy Self-Distillation for Long…
…effectively suppress prediction variance at a fractional computational cost. Empirical results demonstrate that Fast-dDrive redefines the speed-accuracy frontier for driving agents. On the WOD-E2E test set, Fast-dDrive achieves…
…A Large-Scale and High-Quality Benchmark for Generative AI Model and Data Card Generation (2026) Beyond Rating: A Comprehensive Evaluation and Benchmark for AI Reviews (2026) NoveltyAgent: Autonomous Novelty Reporting Agent…
…follower components. AI-generated summary Instruction following is a fundamental capability of large language models (LLMs), yet continuously improving this capability remains challenging. Existing methods typically rely either on costly external supervision…
…AI-generated summary This paper proposes a novel approach to address the challenge that pretrained VLA models often fail to effectively improve performance and reduce adaptation costs during standard supervised finetuning (SFT…