Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Ningyu Zhang

Ningyu Zhang

8 indexed papers

Recent (6 mo)
8
With code
0
Influential cites
0
Benchmarked
0

Publications per year

8
26

Top categories

NLP×8ML×7AI×6Multiagent×3Multimedia×2Robotics×1Vision×1

Frequent co-authors

Buqiang Xu3×
Jizhan Fang3×
Xinle Deng3×
Chen Jiang2×
Haofen Wang2×
Feiyu Xiong2×

Research Timeline

2026
Rethinking Memory as Continuously Evolving Connectivity

The paper proposes FluxMem, a novel connectivity-evolving memory framework that models memory as a dynamic graph to improve LLM agent performance in complex, changing environments.

MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems

The paper introduces MemTrace, a framework that treats LLM memory pipelines as traceable graphs to systematically diagnose and automatically correct memory-related errors, boosting performance by up to 7.62%.

LongDS-Bench: On the Failure of Long-Horizon Agentic Data Analysis

The paper introduces LongDS, a new benchmark for long-horizon, multi-turn data analysis, demonstrating that current AI agents struggle significantly with maintaining and updating complex analytical states over extended interactions.

How LoRA Remembers? A Parametric Memory Law for LLM Finetuning

The paper quantifies the exact parametric memory capacity of LLMs using LoRA and proposes a new optimization strategy, MemFT, to enhance memory fidelity.

ConsisGuard: Aligning Safety Deliberation with Policy Enforcement in LLM Guardrails

The paper introduces ConsisGuard, a framework that addresses the 'deliberation-to-enforcement gap' in LLM guardrails by ensuring that the reasoning process is faithfully and consistently translated into the final safety decision.

LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories

The paper presents a simulation-based workflow and data engine (RoboGenesis) and a vision-language-action model (LabVLA) to enable robots to execute experimental protocols in scientific laboratories.

TokenPilot: Cache-Efficient Context Management for LLM Agents

The paper presents TokenPilot, a dual-granularity context management framework to reduce inference costs in LLM agents.

MemT-1.0: Towards Memory-Native Evolution

This paper introduces MemT-1.0, a new architecture of memory foundation models that internalize memory as part of model computation.

Highlighted terms show continued research focus across papers

Papers

cs.CLcs.LGEmpiricalRecentOct 19, 2026

MemT-1.0: Towards Memory-Native Evolution

Zhiyu Li, Zeyu Zhang, Linyun Liu, Yiwei Bu +70 more

This paper introduces MemT-1.0, a new architecture of memory foundation models that internalize memory as part of model computation.

View →
cs.CLcs.AIcs.LGEmpirical
Recent
Jun 15, 2026

TokenPilot: Cache-Efficient Context Management for LLM Agents

Buqiang Xu, Zirui Xue, Dianmou Chen, Chenyang Fu +11 more

The paper presents TokenPilot, a dual-granularity context management framework to reduce inference costs in LLM agents.

View →
cs.CLcs.AIcs.LGEmpiricalRecentJun 11, 2026

LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories

Baochang Ren, Xinjie Liu, Xi Chen, Yanshuo Liu +14 more

The paper presents a simulation-based workflow and data engine (RoboGenesis) and a vision-language-action model (LabVLA) to enable robots to execute experimental protocols in scientific laboratories.

View →
cs.CLRecentMay 29, 2026

ConsisGuard: Aligning Safety Deliberation with Policy Enforcement in LLM Guardrails

Yan Wang, Zhixuan Chu, Zihao Xue, Zhen Bi +8 more

The paper introduces ConsisGuard, a framework that addresses the 'deliberation-to-enforcement gap' in LLM guardrails by ensuring that the reasoning process is faithfully and consistently translated in…

View →
cs.LGcs.AIcs.CLRecentMay 28, 2026

LongDS-Bench: On the Failure of Long-Horizon Agentic Data Analysis

Kewei Xu, Xiaoben Lu, Shuofei Qiao, Zihan Ding +3 more

The paper introduces LongDS, a new benchmark for long-horizon, multi-turn data analysis, demonstrating that current AI agents struggle significantly with maintaining and updating complex analytical st…

View →
cs.CLcs.AIcs.CVRecentMay 28, 2026

How LoRA Remembers? A Parametric Memory Law for LLM Finetuning

Ziwen Xu, Haiwen Hong, Linsong Yu, Benglei Cui +3 more

The paper quantifies the exact parametric memory capacity of LLMs using LoRA and proposes a new optimization strategy, MemFT, to enhance memory fidelity.

View →
cs.CLcs.AIcs.LGRecentMay 27, 2026

Rethinking Memory as Continuously Evolving Connectivity

Jizhan Fang, Buqiang Xu, Zhixian Wang, Haoliang Cao +11 more

The paper proposes FluxMem, a novel connectivity-evolving memory framework that models memory as a dynamic graph to improve LLM agent performance in complex, changing environments.

View →
cs.CLcs.AIcs.LGRecentMay 27, 2026

MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems

Xinle Deng, Ruobin Zhong, Hujin Peng, Xiaoben Lu +14 more

The paper introduces MemTrace, a framework that treats LLM memory pipelines as traceable graphs to systematically diagnose and automatically correct memory-related errors, boosting performance by up t…

View →