6 papers
Beyond Model Base Retrieval: Weaving Knowledge to Master Fine-grained Neural Network Design
Jialiang Wang, Hanmo Liu, Shimin Di +4
Designing high-performance neural networks for new tasks requires balancing optimization quality with search efficiency. Current methods fail to achieve this balance: neural archit…
Learning to Compose for Cross-domain Agentic Workflow Generation
Jialiang Wang, Shengxiang Xu, Hanmo Liu +5
Automatically generating agentic workflows -- executable operator graphs or codes that orchestrate reasoning, verification, and repair -- has become a practical way to solve comple…
Proficient Graph Neural Network Design by Accumulating Knowledge on Large Language Models
Jialiang Wang, Hanmo Liu, Shimin Di +4
High-level automation is increasingly critical in AI, driven by rapid advances in large language models (LLMs) and AI agents. However, LLMs, despite their general reasoning power,…
RxnNano:Training Compact LLMs for Chemical Reaction and Retrosynthesis Prediction via Hierarchical Curriculum Learning
Ran Li, Shimin Di, Haowei LI +4
Chemical reaction prediction is pivotal for accelerating drug discovery and synthesis planning. Despite advances in data-driven models, current approaches are hindered by an overem…
Beyond path selection: Better LLMs for Scientific Information Extraction with MimicSFT and Relevance and Rule-induced(R)GRPO
Ran Li, Shimin Di, Yuchen Liu +3
Previous study suggest that powerful Large Language Models (LLMs) trained with Reinforcement Learning with Verifiable Rewards (RLVR) only refines reasoning path without improving t…
Learning Towards Emergence: Paving the Way to Induce Emergence by Inhibiting Monosemantic Neurons on Pre-trained Models
Jiachuan Wang, Shimin Di, Tianhao Tang +4
Emergence, the phenomenon of a rapid performance increase once the model scale reaches a threshold, has achieved widespread attention recently. The literature has observed that mon…