17 papers
From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement
Qinsi Wang, Jing Shi, Huazheng Wang +8
Reinforcement Learning with Verifiable Rewards (RLVR) has driven recent progress in reasoning-oriented large language models (LLMs) by enabling large-scale optimization. However, i…
DecodeShare: Tracing the Shared Subspace of LLM Decode-Time Decisions
Zishan Shao, Lixun Zhang, Kangning Cui +10
Large language models (LLMs) handle many tasks with one set of parameters, but under KV-cached inference it is unclear what task-general structure, if any, is used at decode time r…
When No Answer Is Correct: Diagnosing Absent Answer Detection for MLLMs in Video Understanding
Yiheng Wang, Yueqian Lin, Lichen Zhu +3
Multimodal large language models (MLLMs) have made substantial advancements in video understanding, yet the reliability of their responses remains underexplored. This work presents…
MARS: Efficient, Adaptive Co-Scheduling for Heterogeneous Agentic Systems
Yifei Wang, Hancheng Ye, Yechen Xu +8
Large language models (LLMs) are increasingly deployed as the execution core of autonomous agents rather than as standalone text generators. Agentic workloads induce a temporal shi…
Query-Conditioned Evidential Keyframe Sampling for MLLM-Based Long-Form Video Understanding
Yiheng Wang, Lichen Zhu, Yueqian Lin +4
Multimodal Large Language Models (MLLMs) have shown strong performance on video question answering, but their application to long-form videos is constrained by limited context leng…
T2S-Bench & Structure-of-Thought: Benchmarking and Prompting Comprehensive Text-to-Structure Reasoning
Qinsi Wang, Hancheng Ye, Jinhee Kim +12
Think about how human handles complex reading tasks: marking key points, inferring their relationships, and structuring information to guide understanding and responses. Likewise,…
MUSE: A Run-Centric Platform for Multimodal Unified Safety Evaluation of Large Language Models
Zhongxi Wang, Yueqian Lin, Jingyang Zhang +3
Safety evaluation of multimodal large language models requires tracking not only whether an attack succeeds, but also how the interaction unfolds across turns and input modalities.…
Bridging the Perception Gap: A Lightweight Coarse-to-Fine Architecture for Edge Audio Systems
Hengfan Zhang, Yueqian Lin, Hai Helen Li +1
Deploying Audio-Language Models (Audio-LLMs) on edge infrastructure exposes a persistent tension between perception depth and computational efficiency. Lightweight local models ten…
LLaViDA: A Large Language Vision Driving Assistant for Explicit Reasoning and Enhanced Trajectory Planning
Yudong Liu, Spencer Hallyburton, Jiwoo Kim +8
Trajectory planning is a fundamental yet challenging component of autonomous driving. End-to-end planners frequently falter under adverse weather, unpredictable human behavior, or…
KVCOMM: Online Cross-context KV-cache Communication for Efficient LLM-based Multi-agent Systems
Hancheng Ye, Zhengqi Gao, Mingyuan Ma +8
Multi-agent large language model (LLM) systems are increasingly adopted for complex language processing tasks that require communication and coordination among agents. However, the…
AsyncVoice Agent: Real-Time Explanation for LLM Planning and Reasoning
Yueqian Lin, Zhengmian Hu, Jayakumar Subramanian +4
Effective human-AI collaboration on complex reasoning tasks requires that users understand and interact with the model's process, not just receive an output. However, the monolithi…
Voice Evaluation of Reasoning Ability: Diagnosing the Modality-Induced Performance Gap
Yueqian Lin, Zhengmian Hu, Qinsi Wang +6
We present Voice Evaluation of Reasoning Ability (VERA), a benchmark for evaluating reasoning ability in voice-interactive systems under real-time conversational constraints. VERA…
Vision-Zero: Scalable VLM Self-Improvement via Strategic Gamified Self-Play
Qinsi Wang, Bo Liu, Tianyi Zhou +6
Although reinforcement learning (RL) has emerged as a promising approach for improving vision-language models (VLMs) and multimodal large language models (MLLMs), current methods r…
Angles Don't Lie: Unlocking Training-Efficient RL Through the Model's Own Signals
Qinsi Wang, Jinghan Ke, Hancheng Ye +6
Current Reinforcement Fine-tuning (RFT) paradigms for Large Language Models (LLMs) suffer from sample inefficiency due to the redundant exposure of identical queries under uniform…
CoreMatching: A Co-adaptive Sparse Inference Framework with Token and Neuron Pruning for Comprehensive Acceleration of Vision-Language Models
Qinsi Wang, Hancheng Ye, Ming-Yu Chung +6
Vision-Language Models (VLMs) excel across diverse tasks but suffer from high inference costs in time and memory. Token sparsity mitigates inefficiencies in token usage, while neur…
HippoMM: Hippocampal-inspired Multimodal Memory for Long Audiovisual Event Understanding
Yueqian Lin, Jingyang Zhang, Qinsi Wang +5
Comprehending extended audiovisual experiences remains challenging for computational systems, particularly temporal integration and cross-modal associations fundamental to human ep…
Dobi-SVD: Differentiable SVD for LLM Compression and Some New Perspectives
Qinsi Wang, Jinghan Ke, Masayoshi Tomizuka +3
We provide a new LLM-compression solution via SVD, unlocking new possibilities for LLM compression beyond quantization and pruning. We point out that the optimal use of SVD lies in…