17 citations · 64 across the 34 of their papers we have counts for
36 papers
SHarP: Saliency-based Pruning of Agent Harnesses
Xinyi Gao, Qiucheng Wu, Kaizhi Qian +3
Agent harnesses are systems that coordinate model calls, tool use, and task execution to help large language models complete complex tasks. To meet task requirements and address fa…
WorldAuditBench: Interactive 3D World Auditing with Multimodal Agents
Ziyan Jiang, Jingbo Yang, Jiabao Ji +5
As interactive 3D worlds are increasingly used to study intelligent behavior, it becomes important to develop efficient pipelines for identifying anomalies in these simulated envir…
Practical Online KV Cache Compaction for LLM Agents: An Empirical Study
Yujian Liu, Jiabao Ji, Li An +4
LLM agents accumulate long trajectories of reasoning steps, tool calls, and environment feedback, making the KV cache a major inference bottleneck. KV cache compaction can reduce t…
Speculate While You Reason: Teaching Agents to Predict Their Next Tool Call via Joint Agent-Speculator RL
Jiabao Ji, Yujian Liu, Li An +4
Large language model agents often spend substantial wall-clock time waiting for tool call results. Tool-call speculation can hide this latency by predicting and pre-executing an ag…
Unlocking Speech-Text Compositional Powers: Instruction-Following Speech Language Models without Instruction Tuning
Congrui Du, Yang Zhang, Kaizhi Qian +1
Instruction tuning for speech language models (SLMs) is substantially more challenging than for text-based large language models (LLMs), as it requires learning a new modality and…
VISUALSKILL: Multimodal Skills for Computer-Use Agents
Ziyan Jiang, Li An, Yujian Liu +5
Computer-use agents (CUAs) approach human-level performance on standardised benchmarks but still struggle on long-horizon tasks and unseen software. Existing skill libraries addres…
FederatedSkill: Federated Learning for Agentic Skill Evolution
Jingbo Yang, Guanyu Yao, Yang Zhang +3
Modern LLM agents increasingly rely on skill libraries to handle complex tasks, making skill evolution a primary driver of self-improvement. However, isolated single-user task stre…
KT4EQG: Personalized Exercise Question Generation via Knowledge Tracing
Xinyi Gao, Qiucheng Wu, Lu Ding +5
Educational Question Generation (EQG) aims to synthesize customized exercise questions that enhance student learning. An effective EQG system should ideally personalize questions f…
Navigating the Clutter: Waypoint-Based Bi-Level Planning for Multi-Robot Systems
Jiabao Ji, Yongchao Chen, Yang Zhang +4
Multi-robot control in cluttered environments is a challenging problem that involves complex physical constraints, including robot-robot collisions, robot-obstacle collisions, and…
How Well Do Agentic Skills Work in the Wild: Benchmarking LLM Skill Usage in Realistic Settings
Yujian Liu, Jiabao Ji, Li An +3
Agent skills, which are reusable, domain-specific knowledge artifacts, have become a popular mechanism for extending LLM-based agents, yet formally benchmarking skill usage perform…
Learning from Online Videos at Inference Time for Computer-Use Agents
Yujian Liu, Ze Wang, Hao Chen +7
Computer-use agents can operate computers and automate laborious tasks, but despite recent rapid progress, they still lag behind human users, especially when tasks require domain-s…
DEEPAMBIGQA: Ambiguous Multi-hop Questions for Benchmarking LLM Answer Completeness
Jiabao Ji, Min Li, Priyanshu Kumar +2
Large language models (LLMs) with integrated search tools show strong promise in open-domain question answering (QA), yet they often struggle to produce complete answer set to comp…
HarnessLLM: Automatic Testing Harness Generation via Reinforcement Learning
Yujian Liu, Jiabao Ji, Yang Zhang +3
Existing LLM-based automatic test generation methods mainly produce input and expected output pairs to categorize the intended behavior of correct programs. Although straightforwar…
A Reinforcement Learning Framework for Robust and Secure LLM Watermarking
Li An, Yujian Liu, Yepeng Liu +3
Watermarking has emerged as a promising solution for tracing and authenticating text generated by large language models (LLMs). A common approach to LLM watermarking is to construc…
Rethinking the Text-Vision Reasoning Imbalance in MLLMs through the Lens of Training Recipes
Guanyu Yao, Qiucheng Wu, Yang Zhang +3
Multimodal large language models (MLLMs) have demonstrated strong capabilities on vision-and-language tasks. However, recent findings reveal an imbalance in their reasoning capabil…
Collision- and Reachability-Aware Multi-Robot Control with Grounded LLM Planners
Jiabao Ji, Yongchao Chen, Yang Zhang +4
Large language models (LLMs) have demonstrated strong performance in various robot control tasks. However, their deployment in real-world applications remains constrained. Even sta…
Defending LLM Watermarking Against Spoofing Attacks with Contrastive Representation Learning
Li An, Yujian Liu, Yepeng Liu +3
Watermarking has emerged as a promising technique for detecting texts generated by LLMs. Current research has primarily focused on three design criteria: high quality of the waterm…
ThinkPrune: Pruning Long Chain-of-Thought of LLMs via Reinforcement Learning
Bairu Hou, Yang Zhang, Jiabao Ji +4
We present ThinkPrune, a simple yet effective method for pruning the thinking length for long-thinking LLMs, which has been found to often produce inefficient and redundant thinkin…
Instruction-Following Pruning for Large Language Models
Bairu Hou, Qibin Chen, Jianyu Wang +6
With the rapid scaling of large language models (LLMs), structured pruning has become a widely used technique to learn efficient, smaller models from larger ones, delivering superi…
Fictitious Synthetic Data Can Improve LLM Factuality via Prerequisite Learning
Yujian Liu, Shiyu Chang, Tommi Jaakkola +1
Recent studies have identified one aggravating factor of LLM hallucinations as the knowledge inconsistency between pre-training and fine-tuning, where unfamiliar fine-tuning data m…
VSP: Assessing the dual challenges of perception and reasoning in spatial planning tasks for VLMs
Qiucheng Wu, Handong Zhao, Michael Saxon +4
Vision language models (VLMs) are an exciting emerging class of language models (LMs) that have merged classic LM capabilities with those of image processing systems. However, the…
Revisiting Who's Harry Potter: Towards Targeted Unlearning from a Causal Intervention Perspective
Yujian Liu, Yang Zhang, Tommi Jaakkola +1
This paper investigates Who's Harry Potter (WHP), a pioneering yet insufficiently understood method for LLM unlearning. We explore it in two steps. First, we introduce a new task o…
Reversing the Forget-Retain Objectives: An Efficient LLM Unlearning Framework from Logit Difference
Jiabao Ji, Yujian Liu, Yang Zhang +4
As Large Language Models (LLMs) demonstrate extensive capability in learning from documents, LLM unlearning becomes an increasingly important research area to address concerns of L…
A Probabilistic Framework for LLM Hallucination Detection via Belief Tree Propagation
Bairu Hou, Yang Zhang, Jacob Andreas +1
This paper focuses on the task of hallucination detection, which aims to determine the truthfulness of LLM-generated statements. To address this problem, a popular class of methods…