2 citations · 3 across the 17 of their papers we have counts for
18 papers
MMPostTrainBench: Benchmarking Autonomous Research for Multimodal Post-Training
Yuxin Liu, Yuxuan Wang, Zhenxin Lei +9
Autonomous research seeks sustained model improvements through iterative experimentation and feedback. LLM agents show promise in automating machine learning and language-model pos…
Keep It InMind: Benchmarking the Implicit-Association Blind Spot in Agent Memory
Ruizhe Li, Mingxuan Du, Benfeng Xu +1
Long-term memory systems store what a user says in an external store and retrieve it when a related query arrives. This interface rests on an assumption so natural that it is rarel…
Asuka-Bench: Benchmarking Code Agents on Underspecified User Intent and Multi-Round Refinement
Xin Wang, Liangtai Sun, Yaoming Zhu +8
Existing code-generation benchmarks score a single mapping from a complete prompt to a one-shot output. However, real web development is different. Users seldom write a full spec a…
CDRRM: Contrast-Driven Rubric Generation for Reliable and Interpretable Reward Modeling
Dengcan Liu, Fengkai Yang, Xiaohan Wang +7
Reward modeling is essential for aligning Large Language Models(LLMs) with human preferences, yet conventional reward models suffer from poor interpretability and heavy reliance on…
A-RAG: Scaling Agentic Retrieval-Augmented Generation via Hierarchical Retrieval Interfaces
Mingxuan Du, Benfeng Xu, Chiwei Zhu +4
Frontier language models have demonstrated strong reasoning and long-horizon tool-use capabilities. However, existing RAG systems fail to leverage these capabilities. They still re…
WildGraphBench: Benchmarking GraphRAG with Wild-Source Corpora
Pengyu Wang, Benfeng Xu, Licheng Zhang +4
Graph-based Retrieval-Augmented Generation (GraphRAG) organizes external knowledge as a hierarchical graph, enabling efficient retrieval and aggregation of scattered evidence acros…
Wiki Live Challenge: Challenging Deep Research Agents with Expert-Level Wikipedia Articles
Shaohan Wang, Benfeng Xu, Licheng Zhang +5
Deep Research Agents (DRAs) have demonstrated remarkable capabilities in autonomous information retrieval and report generation, showing great potential to assist humans in complex…
FS-Researcher: Test-Time Scaling for Long-Horizon Research Tasks with File-System-Based Agents
Chiwei Zhu, Benfeng Xu, Mingxuan Du +4
Deep research is emerging as a representative long-horizon task for large language model (LLM) agents. However, long trajectories in deep research often exceed model context limits…
DeepResearch Bench II: Diagnosing Deep Research Agents via Rubrics from Expert Reports
Ruizhe Li, Mingxuan Du, Benfeng Xu +3
Deep Research Agents (DRA) aim to help users search the web, synthesize information, and deliver comprehensive investigative reports. Prior benchmarks often either under-evaluate a…
An Index-based Approach for Efficient and Effective Web Content Extraction
Yihan Chen, Benfeng Xu, Xiaorui Wang +1
As web agents (e.g., Deep Research) routinely consume massive volumes of web pages to gather and analyze information, LLM context management -- under large token budgets and low si…
SparseRM: A Lightweight Preference Modeling with Sparse Autoencoder
Dengcan Liu, Jiahao Li, Zheren Fu +4
Reward models (RMs) are a core component in the post-training of large language models (LLMs), serving as proxies for human preference evaluation and guiding model alignment. Howev…
MCP-AgentBench: Evaluating Real-World Language Agent Performance with MCP-Mediated Tools
Zikang Guo, Benfeng Xu, Chiwei Zhu +3
The Model Context Protocol (MCP) is rapidly emerging as a pivotal open standard, designed to enhance agent-tool integration and interoperability, and is positioned to unlock a new…
DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents
Mingxuan Du, Benfeng Xu, Chiwei Zhu +2
Deep Research Agents are a prominent category of LLM-based agents. By autonomously orchestrating multistep web exploration, targeted retrieval, and higher-order synthesis, they tra…
MIRROR: Multi-agent Intra- and Inter-Reflection for Optimized Reasoning in Tool Learning
Zikang Guo, Benfeng Xu, Xiaorui Wang +1
Complex tasks involving tool integration pose significant challenges for Large Language Models (LLMs), leading to the emergence of multi-agent workflows as a promising solution. Re…
From Real to Synthetic: Synthesizing Millions of Diversified and Complicated User Instructions with Attributed Grounding
Chiwei Zhu, Benfeng Xu, Xiaorui Wang +1
The pursuit of diverse, complex, and large-scale instruction data is crucial for automatically aligning large language models (LLMs). While there are methods capable of generating…
Rationales Are Not Silver Bullets: Measuring the Impact of Rationales on Model Performance and Reliability
Chiwei Zhu, Benfeng Xu, An Yang +4
Training language models with rationales augmentation has been shown to be beneficial in many existing works. In this paper, we identify that such a prevailing view does not hold c…
Training LLM-Based Agents with Synthetic Self-Reflected Trajectories and Partial Masking
Yihan Chen, Benfeng Xu, Xiaorui Wang +2
Autonomous agents, which perceive environments and take actions to achieve goals, have become increasingly feasible with the advancements in large language models (LLMs). However,…
Automated Creativity Evaluation for Large Language Models: A Reference-Based Approach
Ruizhe Li, Chiwei Zhu, Benfeng Xu +2
Creative writing is a key capability of Large Language Models (LLMs), with potential applications in literature, storytelling, and various creative domains. However, evaluating the…