7 papers
Atria Dawn: The Dawn of Agentic Superintelligence
Honglin Guo, Tao Gui, Yicheng Chen +139
As AI agents become participants in the development of their successors, they reshape both the production of intelligence and the role of human researchers. We introduce Atria Dawn…
Beyond Human-Likeness: Mapping the Scientific Critique Profiles of LLMs and Human Reviewers
Yunhan Yang, Mike Thelwall, Guoxiu He
Large language models (LLMs) are increasingly discussed as tools for peer review, but their value is often assessed through human-likeness, perceived usefulness, or textual overlap…
Internalizing Academic Writing Workflows for Introduction Generation via Struct-Aware Policy Learning
Meicong Zhang, Tiancheng Su, Jiahao Cheng +3
Generating a rigorous paper introduction with large language models (LLMs) remains challenging, since it requires coordinating background, gap identification, method and contributi…
GraphReview: Scientific Paper Evaluation via LLM-based Graph Evidence Expansion
Pujun Zheng, Wanying Ren, Jiacheng Yao +2
Scientific paper evaluation often involves not only assessing a manuscript itself, but also relating it to contemporaneous research and prior literature. However, existing LLM-base…
Eliminating Agentic Workflow for Introduction Generation with Parametric Stage Tokens
Meicong Zhang, Tiancheng su, Guoxiu He
In recent years, using predefined agentic workflows to guide large language models (LLMs) for literature classification and review has become a research focus. However, writing res…
Entropy-Aware Token Rejection for Improving Speculative Decoding
Tiancheng Su, Meicong Zhang, Guoxiu He
Speculative decoding (SD) accelerates large language model (LLM) inference by using a lightweight draft model to propose tokens and a stronger target model to verify them. However,…
Dynamic Chunking and Selection for Reading Comprehension of Ultra-Long Context in Large Language Models
Boheng Sheng, Jiacheng Yao, Meicong Zhang +1
Large language models (LLMs) often struggle to accurately read and comprehend extremely long texts. Current methods for improvement typically rely on splitting long contexts into f…