long-horizon - 技术专题

GLM-5: From Vibe Coding to Agentic Engineering

The long-horizon computer-use harness. Run AI agents across desktop apps and the CLI for extended periods while preserving task state and making reliable progress on complex workflows. Features fresh-context execution, durable verified state, independent auditing, recoverable progress, and native Claude Code / Codex / OpenClaw integration.

eLLM: Run Long-Horizon Inference Faster on CPUs Than on GPUs

Programmatic memory for long-horizon LLM agents: the harness appends everything to one log, and the agent searches it with code. 97.4% on ARC-AGI-3 (arXiv:2607.20064)

High-performance AI agent for long-horizon tasks and local models. Built on empirical research. 200k+ tokens of work inside a 64k context window. Run a full dev team in 5GB VRAM.

Fixing GRPO training collapse in long-horizon multi-tool agents. A lightweight PRM-Lite + LATA joint approach achieves +37% over vanilla GRPO on τ-bench airline (50-task, multi-turn).

MerchantBench is a 365-day, order-level benchmark for evaluating the long-term coherence of LLM agents in seller-side e-commerce operations.

SWE-Marathon: an ultra long-horizon SWE benchmark

🧠 Awesome Memory-VLA: A curated list of Visual-Language-Action models with memory

Code that accompanies the paper release for "LLMs Corrupt Your Documents When You Delegate"

Rethinking World-Action Model for Compositional and In-context Robotic Manipulation

Official Repository for our paper: PlanBench-XL: Evaluating Long-Horizon Planning of LLM Tool-Use Agents in Large-Scale Tool Ecosystems

OneDayAgent: Towards a Long-Horizon Harness for Autonomous Agents

Code for Scalable Offline Model-Based RL with Action chunking

Simple Long Horizon Agent - A simple yet effective AI agent for learning, experimentation, and long horizon work.

An inference-time framework for long-horizon LLM agents with explicit belief states, consistency validation, and trapping-aware recovery.

MobileMem: Learning from a Year of Mobile Experiences

VLM-RL Hierarchical Loco-Manipulation For Long-Horizon Tasks With G1 robot in Isaac Lab/Sim

Code for Tackling Long-Horizon Tasks with Model-based Offline Reinforcement Learning

[NeurIPS'26] VSearcher: Long-Horizon Multimodal Search Agent via Reinforcement Learning