10 papers
Words to World: Benchmarking LLMs in Interactive Physical Environments
Xinnan Dai, Kaiqi Yang, Yaxin Li +9
Words are increasingly becoming an interface for action: large language models (LLMs) can now invoke tools, operate devices, and make decisions that affect the physical world. Howe…
Magnifying What Matters: Attention-Guided Adaptive Rendering for Visual Text Comprehension
Shenglai Zeng, Qirui Wang, Kai Guo +3
Visual Text Comprehension (VTC) renders text into images for a vision-language model (VLM) to read, sidestepping LLM context-window limits and powering applications from long-page…
Why Retrieval-Augmented Generation Fails: A Graph Perspective
Kai Guo, Xinnan Dai, Zhibo Zhang +5
Retrieval-Augmented Generation (RAG) has become a powerful and widely used approach for improving large language models by grounding generation in retrieved evidence. However, RAG…
When Do Hallucinations Arise? A Graph Perspective on the Evolution of Path Reuse and Path Compression
Xinnan Dai, Kai Yang, Cheng Luo +3
Reasoning hallucinations in large language models (LLMs) often appear as fluent yet unsupported conclusions that violate either the given context or underlying factual knowledge. A…
Fix Before Search: Benchmarking Agentic Query Visual Pre-processing in Multimodal Retrieval-augmented Generation
Jiankun Zhang, Shenglai Zeng, Kai Guo +4
Multimodal Retrieval-Augmented Generation (MRAG) has emerged as a key paradigm for grounding MLLMs with external knowledge. While query pre-processing (e.g., rewriting) is standard…
Attn-GS: Attention-Guided Context Compression for Efficient Personalized LLMs
Shenglai Zeng, Tianqi Zheng, Chuan Tian +10
Personalizing large language models (LLMs) to individual users requires incorporating extensive interaction histories and profiles, but input token constraints make this impractica…
GraphGhost: Tracing Structures Behind Large Language Models
Xinnan Dai, Xianxuan Long, Chung-Hsiang Lo +4
Large Language Models (LLMs) exhibit strong reasoning capabilities on structured tasks, yet the internal mechanisms underlying such behaviors remain poorly understood. Existing int…
Beyond Static Retrieval: Opportunities and Pitfalls of Iterative Retrieval in GraphRAG
Kai Guo, Xinnan Dai, Shenglai Zeng +4
Retrieval-augmented generation (RAG) is a powerful paradigm for improving large language models (LLMs) on knowledge-intensive question answering. Graph-based RAG (GraphRAG) leverag…
Uncovering Graph Reasoning in Decoder-only Transformers with Circuit Tracing
Xinnan Dai, Chung-Hsiang Lo, Kai Guo +3
Transformer-based LLMs demonstrate strong performance on graph reasoning tasks, yet their internal mechanisms remain underexplored. To uncover these reasoning process mechanisms in…
From Sequence to Structure: Uncovering Substructure Reasoning in Transformers
Xinnan Dai, Kai Yang, Jay Revolinsky +4
Recent studies suggest that large language models (LLMs) possess the capability to solve graph reasoning tasks. Notably, even when graph structures are embedded within textual desc…