The Year in AI Papers
The year/Topics/Agent training and self-evolution

Research collection

Agent training and self-evolution

Training agents: environment synthesis, self-evolving and self-improving agent loops, agent RL frameworks, and agent post-training recipes.

Papers
75
Research labs
6
Official code
56

1–50 of 75 papers in this collection

01

Independent research

LongHorizon-Harness: Advancing Long-Horizon Agents for Real-World Tasks

LongHorizon-Harness, from Alibaba's DreamX Team, addresses long-horizon LLM agent failures by reformulating execution as task-state management. It uses a Manage-Execute-Audit (MEA) loop: a manager maintains explicit task state and defines subtasks, a fresh-context executor performs each subtask, and a read-only auditor independently verifies environment…

Ziyu Ma, Hailang Huang, Shun Zou, Yong Wang, et al.
Published
Aug 2026
Citations
0
Code
311 stars
Read summary Original
02

Independent research

Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering

The paper introduces OpenMLE, an open full-stack system for studying recursive self-improvement (RSI) in machine learning engineering (MLE), and Frontis-MA1-35B, a meta-evolution agent trained on this stack. OpenMLE comprises OpenMLE-Gym (5,758 quality-gated executable tasks with sandboxed execution), OpenMLE-ERL (execution-grounded SFT and RL training of…

Junlin Yang, Che Jiang, Yu Fu, Tianwei Luo, et al.
Published
Jul 2026
Citations
0
Code
212 stars
Read summary Original
03

Independent research

Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents

Qwen-UI-Agent is a foundation GUI agent designed for real-world use across mobile, computer, web, and DeepSearch environments. It integrates scalable sandbox environments with a real-device mobile runtime (over 100 devices, 150+ apps), a unified action space interleaving GUI, CLI, and batched actions, an AutoResearch-style data flywheel, and a harness…

Hanzhang Zhou, Panrong Tong, Xu Zhang, Quyu Kong, et al.
Published
Jul 2026
Citations
0
Code
Not linked
Read summary Original
04

Independent research

JarvisHub: An Open Harness for Canvas-Native Multimodal Creative Agents

JarvisHub is an open, canvas-native agent harness designed for long-horizon multimodal creative tasks, addressing the limitations of prompt-to-output tools, chat-based agents, and node-based workflows that fail to maintain a unified project state. It treats an editable canvas as the shared workspace, external memory, and action space, representing…

Yunlong Lin, Zixu Lin, Zhaohu Xing, Biqiang Li, et al.
Published
Jul 2026
Citations
0
Code
277 stars
Read summary Original
05

Independent research

UI-MOPD: Multi-Platform On-Policy Distillation for Continual GUI Agent Learning

The paper introduces UI-MOPD, a method for continual learning of multi-platform GUI agents, addressing challenges like scarce cross-platform data and behavioral pattern mixing. It constructs Uni-GUI, a dataset of ~10K high-quality trajectories from desktop and mobile environments, and proposes multi-teacher on-policy distillation (MOPD) with…

Niu Lian, Alan Chen, Zhehao Yu, Chengzhen Duan, et al.
Published
Jul 2026
Citations
1
Code
56 stars
Read summary Original
06

Independent research

Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent

Agents-A1 is a 35B Mixture-of-Experts agentic model that achieves trillion-parameter-level performance by scaling the agent horizon rather than parameters. The authors built a long-horizon knowledge-action infrastructure that connects external knowledge, actions, observations, and verifier outcomes, producing agentic trajectories averaging 45K tokens.…

Lei Bai, Zongsheng Cao, Yang Chen, Zhiyao Cui, et al.
Published
Jun 2026
Citations
3
Code
529 stars
Read summary Original
07

Independent research

OPID: On-Policy Skill Distillation for Agentic Reinforcement Learning

OPID (On-Policy Skill Distillation) is a framework for agentic reinforcement learning that extracts hierarchical hindsight skills from completed on-policy trajectories to provide dense token-level supervision, complementing sparse outcome-based RL. It represents trajectory hindsight as episode-level skills (global workflows or failure-avoidance rules) and…

Shuo Yang, Jinyang Wu, Zhengxi Lu, Yuhao Shen, et al.
Published
Jun 2026
Citations
8
Code
106 stars
Read summary Original
08

Qwen

Qwen-AgentWorld: Language World Models for General Agents

Qwen-AgentWorld introduces the first language world models (LWMs) for simulating agentic environments across seven domains (MCP, Search, Terminal, SWE, Android, Web, OS). Trained on over 10 million real-world interaction trajectories via a three-stage pipeline (CPT, SFT, RL), the models (35B-A3B and 397B-A17B) predict next environment states given actions.…

Yuxin Zuo, Zikai Xiao, Li Sheng, Fei Huang, et al.
Published
Jun 2026
Citations
0
Code
931 stars
Read summary Original
09

Independent research

Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application

This paper surveys the field of agentic environment engineering for large language models (LLMs), covering the full lifecycle of environment modeling, synthesis, evaluation, and application. It defines agentic environments as dynamic, interactive systems for training and evaluating LLM agents, contrasting them with traditional RL simulators. The survey…

Jiachun Li, Zhuoran Jin, Tianyi Men, Yupu Hao, et al.
Published
Jun 2026
Citations
1
Code
Not linked
Read summary Original
10

Independent research

JoyAI-VL-Interaction: Real-Time Vision-Language Interaction Intelligence

JoyAI-VL-Interaction introduces a paradigm shift from turn-based to proactive, streaming interaction for vision-language models. The authors release an 8B-scale, vision-first model that continuously watches live video and decides each second to stay silent, respond, or delegate complex tasks to a background model. Built on JoyAI-VL 1.0 with AdaCodec for…

Dingyu Yao, Junhao Zhou, Chenxu Yang, Chuanyu Qin, et al.
Published
Jun 2026
Citations
2
Code
1.7K stars
Read summary Original
11

Independent research

Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolution

Role-Agent is a framework that enables bootstrapped agent-environment co-evolution using a single LLM in dual roles. It consists of World-In-Agent (WIA), where the LLM predicts future states after actions, using the alignment between predicted and actual states as a process reward to encourage environment-aware reasoning, and Agent-In-World (AIW), where…

Xucong Wang, Ziyu Ma, Shidong Yang, Tongwen Huang, et al.
Published
Jun 2026
Citations
0
Code
79 stars
Read summary Original
12

MiniMax

The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence

The MiniMax-M2 series is a family of Mixture-of-Experts language models designed to achieve frontier performance with minimal activated parameters. The flagship M2 has 229.9B total parameters but only 9.8B activated per token, using a 62-layer decoder-only Transformer with 256 fine-grained experts, sigmoid gating, full multi-head attention with GQA, a…

Aili Chen, Aonian Li, Baichuan Zhou, Bangwei Gong, et al.
Published
May 2026
Citations
23
Code
Not linked
Read summary Original
13

arXiv.org

Macaron-A2UI: A Model for Generative UI in Personal Agents

The paper introduces Macaron-A2UI, a model for Generative UI in personal agents, which dynamically synthesizes executable UI actions alongside natural language to improve interaction efficiency. The authors build a large-scale corpus of over 14,000 samples from four dialogue sources (MultiWOZ, SGD, ESConv, AnnoMI) using a hybrid rule-and-LLM pipeline with…

Fancy Kong, Congjie Zheng, Murphy Zhuang, Rio Yang, et al.
Published
May 2026
Citations
1
Code
Not linked
Read summary Original
14

arXiv.org

Foundation Protocol: A Coordination Layer for Agentic Society

The Foundation Protocol (FP) is a graph-first coordination layer for an emerging human-AI society, designed to unify heterogeneous entities (agents, tools, humans, institutions, organizations) and support multi-party organization, event-based collaboration, economic primitives (metering, receipts, settlement), and first-class policy, provenance, and audit.…

Bang Liu, Yongfeng Gu, Jiayi Zhang, Zhaoyang Yu, et al.
Published
May 2026
Citations
0
Code
52 stars
Read summary Original
15

arXiv.org

Self-Distilled Agentic Reinforcement Learning

The paper introduces SDAR (Self-Distilled Agentic Reinforcement Learning), a method for post-training multi-turn LLM agents. It addresses issues with On-Policy Self-Distillation (OPSD), which provides dense token-level guidance but suffers from multi-turn instability and asymmetric trust in privileged guidance (e.g., retrieved skills). SDAR keeps GRPO as…

Zhengxi Lu, Zhiyuan Yao, Zhuowen Han, Zi-Han Wang, et al.
Published
May 2026
Citations
13
Code
328 stars
Read summary Original
16

arXiv.org

Video2GUI: Synthesizing Large-Scale Interaction Trajectories for Generalized GUI Agent Pretraining

The paper introduces Video2GUI, a fully automated framework that extracts grounded GUI interaction trajectories from unlabeled internet videos to address the scarcity of large-scale training data for GUI agents. The pipeline uses a coarse-to-fine filtering strategy: first, metadata-based classification (using a fine-tuned Qwen2.5-7B) reduces 500 million…

Weimin Xiong, Shuhao Gu, Bowen Ye, Zihao Yue, et al.
Published
May 2026
Citations
2
Code
38 stars
Read summary Original
17

arXiv.org

MCP-Cosmos: World Model-Augmented Agents for Complex Task Execution in MCP Environments

MCP-Cosmos is a framework that integrates generative World Models (WM) into the Model Context Protocol (MCP) ecosystem to enable predictive task automation. It uses a 'Bring Your Own World Model' (BYOWM) strategy, allowing agents to simulate state transitions and refine plans in a latent space before execution. The framework was evaluated on 24 MCP-Bench…

Giridhar Ganapavarapu, Dhaval Patel
Published
May 2026
Citations
1
Code
Not linked
Read summary Original
18

arXiv.org

ClawGym: A Scalable Framework for Building Effective Claw Agents

ClawGym is a scalable framework for developing Claw-style personal agents, addressing the lack of systematic data synthesis, training, and evaluation in such environments. It introduces ClawGym-SynData, a dataset of 13.5K filtered tasks generated via persona-driven top-down and skill-grounded bottom-up pipelines, with realistic mock workspaces and hybrid…

Fei Bai, Huatong Song, Shuang Sun, Daixuan Cheng, et al.
Published
Apr 2026
Citations
3
Code
Not linked
Read summary Original
19

arXiv.org

From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company

The paper introduces OneManCompany (OMC), a framework that elevates multi-agent systems to an organizational level by decoupling agent capabilities from organizational structure. OMC encapsulates skills, tools, and configurations into portable 'Talents' that run on heterogeneous backends via 'Containers', with six typed organizational interfaces. A…

Zhengxu Yu, Yu Fu, Zhiyuan He, Yuxuan Huang, et al.
Published
Apr 2026
Citations
0
Code
376 stars
Read summary Original
20

arXiv.org

Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence

Agent-World is a self-evolving training arena for general-purpose AI agents, developed by Renmin University of China and ByteDance Seed. It addresses two key bottlenecks in agent training: scalable realism in environment synthesis and continuous self-improvement. The system has two main components: (1) Agentic Environment-Task Discovery, which autonomously…

Guanting Dong, Junting Lu, Junjie Huang, Wanjun Zhong, et al.
Published
Apr 2026
Citations
14
Code
Not linked
Read summary Original
21

arXiv.org

SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization

SKILL0 is a reinforcement learning framework that internalizes agent skills into model parameters, eliminating the need for inference-time skill retrieval. It uses in-context RL during training, where skills are provided as visual context and progressively withdrawn via a Dynamic Curriculum that evaluates each skill's on-policy helpfulness. Experiments on…

Zhengxi Lu, Zhiyuan Yao, Jinyang Wu, Chengcheng Han, et al.
Published
Apr 2026
Citations
47
Code
360 stars
Read summary Original
22

arXiv.org

CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery

CORAL is a framework for autonomous multi-agent evolution on open-ended problems, replacing fixed evolutionary search heuristics with long-running LLM agents that control retrieval, proposal, evaluation, and knowledge accumulation. It uses shared persistent memory (attempts, notes, skills), asynchronous multi-agent execution, and heartbeat-based…

Ao Qu, Han Zheng, Zijian Zhou, Yihao Yan, et al.
Published
Apr 2026
Citations
29
Code
873 stars
Read summary Original
23

arXiv.org

From Static Templates to Dynamic Runtime Graphs: A Survey of Workflow Optimization for LLM Agents

This survey reviews methods for optimizing workflows in large language model (LLM)-based agentic systems, which are modeled as agentic computation graphs (ACGs). The authors propose a taxonomy based on when workflow structure is determined, distinguishing static methods (fixed reusable templates optimized offline) from dynamic methods (structure selected,…

Ling Yue, Kushal Raj Bhandari, Ching-Yun Ko, Dhaval Patel, et al.
Published
Mar 2026
Citations
14
Code
73 stars
Read summary Original
24

arXiv.org

MetaClaw: Just Talk -- An Agent That Meta-Learns and Evolves in the Wild

MetaClaw is a continual meta-learning framework that enables deployed LLM agents to evolve through two complementary mechanisms: skill-driven fast adaptation and opportunistic policy optimization. Skill-driven fast adaptation analyzes failure trajectories to synthesize new behavioral instructions, which are immediately injected into the agent's prompt…

Peng Xia, Jianwen Chen, Xinyu Yang, Haoqin Tu, et al.
Published
Mar 2026
Citations
20
Code
3.5K stars
Read summary Original
25

arXiv.org

Online Experiential Learning for Language Models

The paper introduces Online Experiential Learning (OEL), a framework for large language models to continuously improve from their own deployment experience without human annotations, reward models, or access to user-side environments. OEL operates in two iterated stages: first, transferable experiential knowledge is extracted and accumulated from user-side…

Tianzhu Ye, Li Dong, Qingxiu Dong, Xun Wu, et al.
Published
Mar 2026
Citations
20
Code
Not linked
Read summary Original
26

arXiv.org

OpenClaw-RL: Train Any Agent Simply by Talking

OpenClaw-RL is a framework that enables online reinforcement learning for LLM-based agents by extracting training signals from the next-state signals that follow each agent action (e.g., user replies, tool outputs, GUI changes). It uses a server–client architecture where the RL server hosts the policy behind an inference API and user terminals stream…

Yinjie Wang, Xuyang Chen, Xiaolong Jin, Mengdi Wang, et al.
Published
Mar 2026
Citations
66
Code
5.6K stars
Read summary Original
27

arXiv.org

UI-Venus-1.5 Technical Report

UI-Venus-1.5 is a unified, end-to-end GUI agent family (2B, 8B, and 30B-A3B variants) built on Qwen3-VL, designed for robust real-world applications. It introduces three key advances over UI-Venus-1.0: a Mid-Training stage using 10B tokens from 30+ datasets to establish foundational GUI semantics; Online Reinforcement Learning with full-trajectory rollouts…

Venus Team, Changlong Gao, Zhangxuan Gu, Yulin Liu, et al.
Published
Feb 2026
Citations
27
Code
1K stars
Read summary Original
28

arXiv.org

SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning

SKILLRL is a framework for LLM agents that improves policy learning by distilling raw interaction trajectories into a hierarchical skill library (SKILLBANK) and co-evolving it with the agent's policy during reinforcement learning. It uses a teacher model to extract general and task-specific skills from both successful and failed episodes, achieving 10-20x…

Peng Xia, Jianwen Chen, Hanyang Wang, Jiaqi Liu, et al.
Published
Feb 2026
Citations
150
Code
928 stars
Read summary Original
29

arXiv.org

TermiGen: High-Fidelity Environment and Robust Trajectory Synthesis for Terminal Agents

TermiGen is an end-to-end pipeline for synthesizing verifiable terminal environments and robust expert trajectories to train open-weight LLMs for terminal tasks. It addresses two key challenges: scarcity of high-fidelity executable environments and exposure bias from standard expert trajectories. Phase I uses a multi-agent system to generate diverse tasks…

Kaijie Zhu, Yuzhou Nie, Yijiang Li, Yiming Huang, et al.
Published
Feb 2026
Citations
13
Code
73 stars
Read summary Original
30

arXiv.org

Baichuan-M3: Modeling Clinical Inquiry for Reliable Medical Decision-Making

Baichuan-M3 is a medical-enhanced large language model designed to shift from passive question-answering to active, clinical-grade decision support. It addresses limitations in open-ended consultations by modeling the systematic workflow of a physician, with key capabilities including proactive information acquisition, long-horizon reasoning, and adaptive…

Baichuan-M3 Team, :, Chengfeng Dou, Fan Yang, et al.
Published
Feb 2026
Citations
8
Code
254 stars
Read summary Original
31

arXiv.org

QuantaAlpha: An Evolutionary Framework for LLM-Driven Alpha Mining

QuantaAlpha is an evolutionary alpha mining framework that treats each end-to-end mining run as a trajectory and improves factors via trajectory-level mutation and crossover. It addresses limitations in existing agentic systems, such as fragile controllability, limited trustworthiness, and constrained exploration, by localizing suboptimal steps for…

Jun Han, Shuo Zhang, Wei Li, Yifan Dong, et al.
Published
Feb 2026
Citations
7
Code
1.4K stars
Read summary Original
32

arXiv.org

AOrchestra: Automating Sub-Agent Creation for Agentic Orchestration

AORCHESTRA is an agentic framework that automates sub-agent creation for complex, long-horizon tasks. It introduces a unified four-tuple abstraction (Instruction, Context, Tools, Model) to model any agent, enabling on-demand specialization. A central orchestrator decomposes tasks, curates context, selects tools and models, and delegates execution to…

Jianhao Ruan, Zhihao Xu, Yiran Peng, Fashen Ren, et al.
Published
Feb 2026
Citations
18
Code
153 stars
Read summary Original
33

Google DeepMind

MARS: Modular Agent with Reflective Search for Automated AI Research

MARS (Modular Agent with Reflective Search) is a framework for automating AI research, specifically addressing the bottleneck of complex machine learning engineering (MLE) tasks. It uses three pillars: Budget-Aware Planning via cost-constrained Monte Carlo Tree Search (MCTS) to balance performance with execution cost, Modular Construction through a…

Jiefeng Chen, Bhavana Dalvi Mishra, Jaehyun Nam, Rui Meng, et al.
Published
Feb 2026
Citations
12
Code
39 stars
Read summary Original
34

Moonshot AI

Kimi K2.5: Visual Agentic Intelligence

Kimi K2.5 is an open-source multimodal agentic model that jointly optimizes text and vision through techniques including joint pre-training, zero-vision SFT, and joint reinforcement learning. It introduces Agent Swarm, a parallel agent orchestration framework using Parallel-Agent Reinforcement Learning (PARL) to decompose tasks into heterogeneous…

Kimi Team, Tongtong Bai, Yifan Bai, Yiping Bao, et al.
Published
Feb 2026
Citations
313
Code
2.3K stars
Read summary Original
35

arXiv.org

daVinci-Agency: Unlocking Long-Horizon Agency Data-Efficiently

The paper introduces daVinci-Agency, a data synthesis paradigm for training LLM agents on long-horizon tasks. It mines chains of semantically linked GitHub Pull Requests (PRs) to create training trajectories that capture task decomposition, long-term consistency, and iterative refinement. The method constructs task chains from up to five PRs, yielding…

Mohan Jiang, Dayuan Fu, Junhao Shi, Ji Zeng, et al.
Published
Feb 2026
Citations
3
Code
38 stars
Read summary Original
36

arXiv.org

ASTRA: Automated Synthesis of agentic Trajectories and Reinforcement Arenas

ASTRA is a fully automated, end-to-end framework for training tool-augmented language model agents. It combines a trajectory synthesis pipeline that uses the static topology of tool-call graphs to generate diverse, multi-turn tool-use trajectories for supervised fine-tuning (SFT), with an environment synthesis framework that converts decomposed…

Xiaoyu Tian, Haotian Wang, Shuaiting Chen, Hao Zhou, et al.
Published
Jan 2026
Citations
3
Code
151 stars
Read summary Original
37

arXiv.org

LongCat-Flash-Thinking-2601 Technical Report

LongCat-Flash-Thinking-2601 is a 560B-parameter open-source Mixture-of-Experts (MoE) reasoning model with 27B activated parameters, achieving state-of-the-art performance among open-source models on agentic benchmarks. Its training combines a unified framework with domain-parallel expert training, environment scaling, and a robust RL pipeline. Key…

Meituan LongCat Team, Anchun Gui, Bei Li, Bingyang Tao, et al.
Published
Jan 2026
Citations
23
Code
259 stars
Read summary Original
38

arXiv.org

EvoCUA: Evolving Computer Use Agents via Learning from Scalable Synthetic Experience

EvoCUA is a native computer-use agent that replaces static imitation with a self-sustaining evolutionary cycle integrating verifiable data synthesis, scalable interaction infrastructure, and iterative policy optimization. A synthesis engine generates diverse tasks with executable validators, while a high-throughput sandbox platform orchestrates tens of…

Taofeng Xue, Chong Peng, Mianqiu Huang, Linsen Guo, et al.
Published
Jan 2026
Citations
31
Code
336 stars
Read summary Original
39

arXiv.org

Agentic Reasoning for Large Language Models

This survey introduces agentic reasoning, a paradigm shift where large language models (LLMs) act as autonomous agents that plan, act, and learn through continual interaction with their environment, rather than passively generating sequences. The authors organize agentic reasoning along three complementary dimensions: foundational agentic reasoning (core…

Tianxin Wei, Ting-Wei Li, Zhining Liu, Xuying Ning, et al.
Published
Jan 2026
Citations
34
Code
1.3K stars
Read summary Original
40

arXiv.org

Collaborative Multi-Agent Test-Time Reinforcement Learning for Reasoning

MATTRL is a framework that improves multi-agent LLM reasoning at test time by injecting structured textual experience into multi-agent deliberation, avoiding costly weight updates. It forms a team of specialist agents, retrieves relevant experiences from a pool, and reaches consensus. The experience pool is built by scoring utterances with individual and…

Zhiyuan Hu, Yunhai Hu, Juncheng Liu, Shuyue Stella Li, et al.
Published
Jan 2026
Citations
3
Code
Not linked
Read summary Original
41

arXiv.org

User-Oriented Multi-Turn Dialogue Generation with Tool Use at scale

This paper introduces a user-oriented simulation framework for generating multi-turn tool-use dialogue data at scale, addressing limitations of static toolsets and single-shot trajectories. The authors first developed a task-oriented pipeline using an LRM-based simulator to dynamically synthesize tools and tasks, but found it produced minimal-interaction,…

Jungho Cho, Minbyul Jeong, Sungrae Park
Published
Jan 2026
Citations
1
Code
Not linked
Read summary Original
42

Qwen

Let It Flow: Agentic Crafting on Rock and Roll, Building the ROME Model within an Open Agentic Learning Ecosystem

This technical report introduces the Agentic Learning Ecosystem (ALE), a full-stack infrastructure for developing agentic LLMs, and ROME, an open-source agent model trained within it. ALE comprises three components: ROLL, a scalable RL training framework; ROCK, a sandboxed environment execution engine; and iFlow CLI, an agent framework for context…

Weixun Wang, XiaoXiao Xu, Wanhe An, Fangwen Dai, et al.
Published
Dec 2025
Citations
23
Code
3.4K stars
Read summary Original
43

Independent research

Youtu-LLM: Unlocking the Native Agentic Potential for Lightweight Large Language Models

Youtu-LLM is a 1.96B-parameter language model pre-trained from scratch to achieve native agentic intelligence, balancing efficiency with strong reasoning and planning. It uses a dense Multi-Latent Attention (MLA) architecture with a 128k context window and a STEM-oriented tokenizer. Training follows a 'Commonsense-STEM-Agent' curriculum over 10.84T tokens,…

Junru Lu, Jiarui Qin, Lingfeng Qiao, Yinghui Li, et al.
Published
Dec 2025
Citations
5
Code
593 stars
Read summary Original
44

arXiv.org

Youtu-Agent: Scaling Agent Productivity with Automated Generation and Hybrid Policy Optimization

Youtu-Agent is a modular framework addressing high configuration costs and static capabilities in LLM agents. It decouples environments, toolkits, and agents via a YAML-based system, enabling automated generation through a deterministic Workflow mode and a flexible Meta-Agent mode that synthesizes tools, prompts, and configurations. For continuous…

Yuchen Shi, Yuzheng Cai, Siqi Cai, Zihan Xu, et al.
Published
Dec 2025
Citations
5
Code
4.6K stars
Read summary Original
45

arXiv.org

Adaptation of Agentic AI: A Survey of Post-Training, Memory, and Skills

This survey examines how agentic AI systems are adapted after pretraining, organizing the field into a four-paradigm framework based on what is optimized (the agent or its tools) and the source of the adaptation signal (tool execution or agent output). The paradigms are: A1 (tool-execution-signaled agent adaptation), A2 (agent-output-signaled agent…

Pengcheng Jiang, Jiacheng Lin, Zhiyi Shi, Zifeng Wang, et al.
Published
Dec 2025
Citations
15
Code
679 stars
Read summary Original
46

arXiv.org

Step-GUI Technical Report

This technical report introduces Step-GUI, a family of multimodal GUI agent models (4B/8B) built on Qwen3-VL, achieving state-of-the-art performance across benchmarks (8B: 80.2% AndroidWorld, 48.5% OSWorld, 62.6% ScreenShot-Pro). The core innovation is the Calibrated Step Reward System (CSRS), a self-evolving training pipeline that converts model-generated…

Haolong Yan, Jia Wang, Xin Huang, Yeqing Shen, et al.
Published
Dec 2025
Citations
27
Code
Not linked
Read summary Original
47

arXiv.org

Nex-N1: Agentic Models Trained via a Unified Ecosystem for Large-Scale Environment Construction

The paper introduces Nex-N1, a series of agentic models trained using a unified ecosystem (NexAU, NexA4A, NexGAP) for large-scale environment construction. NexAU is a modular runtime for scalable agent frameworks, NexA4A automatically generates diverse agent hierarchies from natural language, and NexGAP generates end-to-end agentic trajectories using real…

Nex-AGI Team, :, Yuxuan Cai, Lu Chen, et al.
Published
Dec 2025
Citations
15
Code
Not linked
Read summary Original
48

arXiv.org

Guided Self-Evolving LLMs with Minimal Human Supervision

The paper introduces R-FEW, a guided self-evolving framework for large language models (LLMs) that uses minimal human supervision to achieve stable and controllable self-improvement. R-FEW addresses issues like concept drift and diversity collapse in unguided self-play methods (e.g., R-Zero) by integrating a few-shot grounded Challenger and an online…

Wenhao Yu, Zhenwen Liang, Chengsong Huang, Kishan Panaganti, et al.
Published
Dec 2025
Citations
31
Code
Not linked
Read summary Original
49

NVIDIA

ToolOrchestra: Elevating Intelligence via Efficient Model and Tool Orchestration

The paper introduces ToolOrchestra, a method for training small language models as orchestrators that coordinate diverse tools, including basic tools (web search, code interpreter), specialized LLMs (math, coding), and generalist LLMs (GPT-5, Claude Opus 4.1). The resulting 8B-parameter Orchestrator model is trained end-to-end with reinforcement learning…

Hongjin Su, Shizhe Diao, Ximing Lu, Mingjie Liu, et al.
Published
Nov 2025
Citations
32
Code
752 stars
Read summary Original
50

arXiv.org

Agent0: Unleashing Self-Evolving Agents from Zero Data via Tool-Integrated Reasoning

Agent0 is a fully autonomous framework that evolves LLM agents from zero external data by combining tool-integrated reasoning with multi-step co-evolution. It initializes two agents from the same base LLM: a curriculum agent trained via GRPO to generate frontier tasks, rewarded by the executor's uncertainty (self-consistency), tool-use frequency, and a…

Peng Xia, Kaide Zeng, Jiaqi Liu, Can Qin, et al.
Published
Nov 2025
Citations
54
Code
1.2K stars
Read summary Original