人工智能学院12篇论文被ICLR’2026接收

发布时间:2026-01-30 点击:

国际学习表征会议(International Conference on Learning Representations,简称ICLR)是由图灵奖得主Yoshua Bengio和Yann LeCun于2013年创立的深度学习领域顶级学术会议。该会议每年举办一次,聚焦表征学习与深度学习的前沿研究,涵盖机器视觉、自然语言处理、计算生物学等应用领域。ICLR’2026将于2026年4月23日至27日‌在巴西里约热内卢的举行。人工智能学院师生共有12篇论文被该会议接收。


论文题目:Imitating the Truth: Attention-aware Truth-Guided Enhancement for Hallucination Mitigation in Large Vision-Language Models

第一作者:任海瑞(2022级博士生)

通讯作者:郭丹丹,常毅


论文题目:LLM as an Algorithmist: Enhancing Anomaly Detectors via Programmatic Synthesis

第一作者:叶航廷(2022级博士生)

通讯作者:郭丹丹,常毅


论文题目:A Guardrail for Safety Preservation: When Safety-Sensitive Subspace Meets Harmful-Resistant Null-Space

第一作者:张丙杰(2024级博士生)

通讯作者:杨一博,郭丹丹


论文题目:RD-HRL: Generating Reliable Sub-Goals for Long-Horizon Sparse-Reward Tasks

第一作者:单怡翔(22级博士生)

通讯作者:龙婷,常毅


论文题目:FaithCoT-Bench: Benchmarking Instance-Level Faithfulness of Chain-of-Thought Reasoning

第一作者:沈旭(23级硕士生)

通讯作者:王鑫


论文题目:BA-LoRA: Bias-Alleviating Low-Rank Adaptation to Mitigate Catastrophic Inheritance in Large Language Models

第一作者:常宇鹏(23级博士生)

通讯作者:邬渊


论文题目:Distribution-Aware Multi-Granularity Phase Coding: Towards Lower Conversion Error for Spike-Driven Large Language Models

主要作者:Hanyuan Zheng, Haozhen Zhang, Tianshuo Chen, Zhaogeng Liu


论文题目:Three Forward, One Backward: Memory-Efficient Full-Rank Fine-Tuning of Large Models via Extra Forward Passes

主要作者:Jia Zhang, Tianshuo Chen, Hualin Zhang, Yu Bai, Zhaogeng Liu


论文题目:Online Black-Box Prompt Optimization with Regret Guarantees under Noisy Feedback

主要作者:Jinjie Fang, Runwen You, Wenkang Wang, Ganyu Wang, Haozhen Zhang


论文题目:LinearRAG: Linear Graph Retrieval Augmented Generation on Large-scale Corpora.

主要作者:Luyao Zhuang, Shengyuan Chen, Yilin Xiao, Huachi Zhou, Yujing Zhang, Hao Chen, Qinggang Zhang, Xiao Huang.


论文题目:When to use Graphs in RAG: A Comprehensive Analysis for Graph Retrieval-Augmented Generation.

主要作者:Zhishang Xiang, Chuanjie Wu, Qinggang Zhang, Shengyuan Chen, Zijin Hong, Xiao Huang, Jinsong Su.


论文题目:Temporal Graph Thumbnail: Robust Representation Learning with Global Evolutionary Skeleton.

主要作者:Weining Shi, Zhisen Wen, Qinggang Zhang, Chentao Zhang, Zhihong Zhang.