YS
Yangqiu Song
cs.AIcs.CLcs.CRcs.SE
On Valency
published · living versionsW_d9x5tsjc·v1 · currentpublished
DAGLoop: Aligning Method Design with Experimental Feedback in Automated AI Research
with Dengyuxing Wang, Yonggang Zhang, Qianshu Cai, Huajiang Zheng +2
1 version
W_78vhtbmk·v1published
W_ssv97bb5·v1published
Preprints & journals
9 papers in the corpus · 2026MOSAIC: An Omni-Modal Benchmark for Long-Term Memory and Cross-Modal Reasoning in Conversational Agents2610.29249v1 · Hong Ting Tsang, Boyi Zhang, Haoyu Huang et al.2026 · 0 citationsarXivon Valency
AdaExploreBench: Benchmarking LLM Agents for Proactive Exploration and Test-time Adaptation in Evolving Environments2610.25950v1 · Rui Wang, Zishen Lai, Jiayu Liu et al.2026 · 0 citationsarXivon Valency
CoworkEnv: Synthesizing Training Environments for Long-Horizon Cowork Agents2610.24960v1 · Zhongwei Xie, Yiming Du, Jiaxin Bai et al.2026 · 0 citationsarXivon Valency
DAGLoop: Aligning Method Design with Experimental Feedback in Automated AI Research2610.25089v1 · Dengyuxing Wang, Yonggang Zhang, Qianshu Cai et al.2026 · 0 citationsarXivon Valency
JockeyBench: Evaluating LLM Agents on Horse-Racing Decision-Making2610.19589v1 · Zhongwei Xie, Xingsu Li, Jingqiao Yu et al.2026 · 0 citationsarXiv
Jev-Guard: Fast Routing and Auditing Agent Actions over Editable Policy Trees2610.18326v1 · Wenbin Hu, Huihao Jing, Haochen Shi et al.2026 · 0 citationsarXiv
Beyond the Verdict: Agentic Code Judging with Verifiable Feedback2610.18760v1 · Zheye Deng, Sha Li, Changlong Yu et al.2026 · 0 citationsarXiv
Same Operation, Different Permission: Evaluating Whether Guardrails Honor Authorization2610.16838v1 · Huihao Jing, Shaojin Chen, Wenbin Hu et al.2026 · 0 citationsarXiv
SkillHorizon: Constructing Long-Horizon Agent Tasks through Skill Composition2610.14966v1 · Yuxuan Liu, Zhaochen Su, Hangyu Guo et al.2026 · 0 citationsarXiv
Career total: 10 works. 9 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.