Cover Image

Mock Worlds, Real Skills: Building Small Agentic Language Models with Synthetic Tasks, Simulated Environments, and Rubric-Based Rewards

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Lyu, Yuanjie, Wang, Chengyu, Shen, Lei, Huang, Jun, Xu, Tong
Format:	Preprint
Published:	2026
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

From Correction to Mastery: Reinforced Distillation of Large Language Model Agents
by: Lyu, Yuanjie, et al.
Published: (2025)

AgenticQwen: Training Small Agentic Language Models with Dual Data Flywheels for Industrial-Scale Tool Use
by: Lyu, Yuanjie, et al.
Published: (2026)

OmniThoughtVis: A Scalable Distillation Pipeline for Deployable Multimodal Reasoning Models
by: Yue, Yuanhao, et al.
Published: (2026)

OpenRubrics: Towards Scalable Synthetic Rubric Generation for Reward Modeling and LLM Alignment
by: Liu, Tianci, et al.
Published: (2025)

Streamlining the Collaborative Chain of Models into A Single Forward Pass in Generation-Based Tasks
by: Lyu, Yuanjie, et al.
Published: (2025)

Rubrics to Tokens: Bridging Response-level Rubrics and Token-level Rewards in Instruction Following Tasks
by: Xu, Tianze, et al.
Published: (2026)

AutoRubric: Rubric-Based Generative Rewards for Faithful Multimodal Reasoning
by: Jia, Mengzhao, et al.
Published: (2025)

AdaRubric: Task-Adaptive Rubrics for Reliable LLM Agent Evaluation and Reward Learning
by: Ding, Liang
Published: (2026)

OccuBench: Evaluating AI Agents on Real-World Professional Tasks via Language Environment Simulation
by: Hu, Xiaomeng, et al.
Published: (2026)

RubricEM: Meta-RL with Rubric-guided Policy Decomposition beyond Verifiable Rewards
by: Li, Gaotang, et al.
Published: (2026)

Bootstrapping Post-training Signals for Open-ended Tasks via Rubric-based Self-play on Pre-training Text
by: Huang, Chengyu, et al.
Published: (2026)

Agent World Model: Infinity Synthetic Environments for Agentic Reinforcement Learning
by: Wang, Zhaoyang, et al.
Published: (2026)

Curing Miracle Steps in LLM Mathematical Reasoning with Rubric Rewards
by: Yuan, Youliang, et al.
Published: (2025)

Xpertbench: Expert Level Tasks with Rubrics-Based Evaluation
by: Liu, Xue, et al.
Published: (2026)

Inducing Programmatic Skills for Agentic Tasks
by: Wang, Zora Zhiruo, et al.
Published: (2025)

Distilling Instruction-following Abilities of Large Language Models with Task-aware Curriculum Planning
by: Yue, Yuanhao, et al.
Published: (2024)

Self-Rewarding Rubric-Based Reinforcement Learning for Open-Ended Reasoning
by: Ye, Zhiling, et al.
Published: (2025)

Code Aesthetics with Agentic Reward Feedback
by: Xiao, Bang, et al.
Published: (2025)

PLawBench: A Rubric-Based Benchmark for Evaluating LLMs in Real-World Legal Practice
by: Shi, Yuzhen, et al.
Published: (2026)

Optimsyn: Influence-Guided Rubrics Optimization for Synthetic Data Generation
by: Fan, Zhiting, et al.
Published: (2026)

Training Language Models to Generate Text with Citations via Fine-grained Rewards
by: Huang, Chengyu, et al.
Published: (2024)

Training AI Co-Scientists Using Rubric Rewards
by: Goel, Shashwat, et al.
Published: (2025)

Chaining the Evidence: Robust Reinforcement Learning for Deep Search Agents with Citation-Aware Rubric Rewards
by: Zhang, Jiajie, et al.
Published: (2026)

OpenReward: Learning to Reward Long-form Agentic Tasks via Reinforcement Learning
by: Hu, Ziyou, et al.
Published: (2025)

Alternating Reinforcement Learning for Rubric-Based Reward Modeling in Non-Verifiable LLM Post-Training
by: Xu, Ran, et al.
Published: (2026)

Rubric-Guided Process Reward for Stepwise Model Routing
by: Ye, Shenghao, et al.
Published: (2026)

Building a Family of Data Augmentation Models for Low-cost LLM Fine-tuning on the Cloud
by: Yue, Yuanhao, et al.
Published: (2024)

DeltaRubric: Generative Multimodal Reward Modeling via Joint Planning and Verification
by: Liu, Rui, et al.
Published: (2026)

EvoRubric: Self-Evolving Rubric-Driven RL for Open-Ended Generation
by: Guan, Xin, et al.
Published: (2026)

SkillLearnBench: Benchmarking Continual Learning Methods for Agent Skill Generation on Real-World Tasks
by: Zhong, Shanshan, et al.
Published: (2026)

MCPVerse: An Expansive, Real-World Benchmark for Agentic Tool Use
by: Lei, Fei, et al.
Published: (2025)

R3: Robust Rubric-Agnostic Reward Models
by: Anugraha, David, et al.
Published: (2025)

Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains
by: Gunjal, Anisha, et al.
Published: (2025)

Configurable Preference Tuning with Rubric-Guided Synthetic Data
by: Gallego, Víctor
Published: (2025)

Building a Foundational Guardrail for General Agentic Systems via Synthetic Data
by: Huang, Yue, et al.
Published: (2025)

Generating Event-oriented Attribution for Movies via Two-Stage Prefix-Enhanced Multimodal LLM
by: Lyu, Yuanjie, et al.
Published: (2024)

Thinking with DistilQwen: A Tale of Four Distilled Reasoning and Reward Model Series
by: Cai, Wenrui, et al.
Published: (2025)

Boosting Reward Model with Preference-Conditional Multi-Aspect Synthetic Data Generation
by: Shen, Jiaming, et al.
Published: (2024)

A Short Survey on Small Reasoning Models: Training, Inference, Applications and Research Directions
by: Wang, Chengyu, et al.
Published: (2025)

TOUCAN: Synthesizing 1.5M Tool-Agentic Data from Real-World MCP Environments
by: Xu, Zhangchen, et al.
Published: (2025)