15 papers
miniReranker: Efficient Multimodal Reranking through Visual Cache Reuse and Interaction Sparsity
Yingqi Fan, Xuan Lu, Anhao Zhao +6
Multimodal large language models (MLLMs) have recently shown strong potential as point-wise rerankers by directly modeling query--document relevance through next-token prediction.…
PowerOPD: Stabilizing On-Policy Distillation with Bounded Power Transformation
Anhao Zhao, Junlong Tong, Yingqi Fan +3
Standard on-policy distillation (OPD) for large language models estimates the reverse-KL objective using student-sampled tokens, yielding an unbiased single-sample Monte Carlo esti…
AdaSR: Adaptive Streaming Reasoning with Hierarchical Relative Policy Optimization
Junlong Tong, Wenqi Xu, Yingqi Fan +4
Large reasoning models typically follow a read-then-think paradigm: they observe the complete input, reason over a static context, and then produce the answer. Yet many real-world…
CompRank: Efficient LLM Reranking via Token-Level Compression and Decoding-Free Scoring
Xuan Lu, Haohang Huang, Yingqi Fan +5
Large language model (LLM) rerankers have become an important component of modern retrieval and retrieval-augmented generation pipelines, but their high computational cost limits t…
ProactiveLLM: Learning Active Interaction for Streaming Large Language Models
Junlong Tong, Yao Zhang, Anhao Zhao +3
Standard Large Language Models (LLMs) follow a read-then-generate paradigm, causing unnecessary latency and computation. Streaming LLMs alleviate this issue by generating while rec…
Decoupling KL and Trajectories: A Unified Perspective for SFT, DAgger, Offline RL, and OPD in LLM Distillation
Anhao Zhao, Haoran Xin, Yingqi Fan +3
Knowledge distillation is central to LLM post-training, yet its design space remains poorly understood, especially alongside reinforcement learning (RL). We show that the prevailin…
What Do Visual Tokens Really Encode? Uncovering Sparsity and Redundancy in Multimodal Large Language Models
Yingqi Fan, Junlong Tong, Anhao Zhao +1
Multimodal large language models (MLLMs) project visual tokens into the embedding space of language models, yet the internal structuring and processing of visual semantics remain p…
HiDrop: Hierarchical Vision Token Reduction in MLLMs via Late Injection, Concave Pyramid Pruning, and Early Exit
Hao Wu, Yingqi Fan, Jinyang Dai +3
The quadratic computational cost of processing vision tokens in Multimodal Large Language Models (MLLMs) hinders their widespread adoption. While progressive vision token pruning o…
On-Policy Supervised Fine-Tuning for Efficient Reasoning
Anhao Zhao, Ziyang Chen, Junlong Tong +6
Large reasoning models (LRMs) are commonly trained with reinforcement learning (RL) to explore long chain-of-thought reasoning, achieving strong performance at high computational c…
ViCA: Efficient Multimodal LLMs with Vision-Only Cross-Attention
Wenjie Liu, Hao Wu, Xin Qiu +6
Modern multimodal large language models (MLLMs) adopt a unified self-attention design that processes visual and textual tokens at every Transformer layer, incurring substantial com…
: Decoding Discontinuous Cross-Modal Dynamics for Efficient Multimodal LLMs
Yingqi Fan, Anhao Zhao, Jinlan Fu +5
Multimodal Large Language Models (MLLMs) have achieved strong performance across vision-language tasks, but suffer from significant computational overhead due to the quadratic grow…
StreamingThinker: Large Language Models Can Think While Reading
Junlong Tong, Yingqi Fan, Anhao Zhao +2
Large language models (LLMs) have demonstrated remarkable capabilities in chain of thought (CoT) reasoning. However, the current LLM reasoning paradigm initiates thinking only afte…
SkipGPT: Dynamic Layer Pruning Reinvented with Token Awareness and Module Decoupling
Anhao Zhao, Fanghua Ye, Yingqi Fan +4
Large language models (LLMs) achieve remarkable performance across tasks but incur substantial computational costs due to their deep, multi-layered architectures. Layer pruning has…
LLM as Effective Streaming Processor: Bridging Streaming-Batch Mismatches with Group Position Encoding
Junlong Tong, Jinlan Fu, Zixuan Lin +4
Large Language Models (LLMs) are primarily designed for batch processing. Existing methods for adapting LLMs to streaming rely either on expensive re-encoding or specialized archit…
Multi-Layer Visual Feature Fusion in Multimodal LLMs: Methods, Analysis, and Best Practices
Junyan Lin, Haoran Chen, Yue Fan +5
Multimodal Large Language Models (MLLMs) have made significant advancements in recent years, with visual features playing an increasingly critical role in enhancing model performan…