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

Peng Wei

5 indexed papers

Recent (6 mo)
3
With code
0
Influential cites
0
Benchmarked
0

Publications per year

5
26

Top categories

ML×2NLP×2Society×2AI×2Crypto×2Robotics×1Multiagent×1Systems and Control×1

Frequent co-authors

Xiaosong Han2×
Ke Chen2×
Xindi Dai2×
Di Liang2×
Minlong Peng2×
Wei Pang2×

Research Timeline

2026
Preserving Decision Sovereignty in Military AI: A Trade-Secret-Safe Architectural Framework for Model Replaceability, Human Authority, and State Control

The paper proposes the Energetic Paradigm, a model-agnostic architectural framework that allows states to maintain decision sovereignty and control over military AI systems, even when using proprietary, commercially sourced analytical models.

A Public Theory of Distillation Resistance via Constraint-Coupled Reasoning Architectures

The paper proposes a theoretical framework, called constraint-coupled reasoning, to make AI models less susceptible to knowledge distillation by coupling high-level capabilities to internal stability constraints.

TRACE: Discovering Task-Specific Parameter via Adaptation-Aware Probing for Continual Fine-Tuning

TRACE proposes a novel method to mitigate catastrophic forgetting in continual LLM fine-tuning by identifying and isolating a small, task-specific subset of essential parameters for each task.

TRACE: Discovering Task-Specific Parameter via Adaptation-Aware Probing for Continual Fine-Tuning

TRACE proposes a novel method to mitigate catastrophic forgetting in continual LLM fine-tuning by identifying and isolating a small, task-specific subset of essential parameters for each task.

Runtime Safety Filtering for Learned Small UAS Separation Policies under GNSS Degradation

This paper evaluates two approaches for maintaining safe separation between small Unmanned Aircraft Systems (sUAS) in urban environments with degraded Global Navigation Satellite System (GNSS) signals: filtering the policy's actions or its observations. The results show that observation filtering significantly reduces mid-air collisions and remains robust to the tradeoff between separation distance and closing rate.

Highlighted terms show continued research focus across papers

Papers

cs.ROcs.LGcs.MAEmpiricalRecentJul 10, 2026

Runtime Safety Filtering for Learned Small UAS Separation Policies under GNSS Degradation

Alex Zongo, Peng Wei

This paper evaluates two approaches for maintaining safe separation between small Unmanned Aircraft Systems (sUAS) in urban environments with degraded Global Navigation Satellite System (GNSS) signals…

View →
cs.CLRecent
May 29, 2026

TRACE: Discovering Task-Specific Parameter via Adaptation-Aware Probing for Continual Fine-Tuning

Xiaosong Han, Ke Chen, Xindi Dai, Di Liang +6 more

TRACE proposes a novel method to mitigate catastrophic forgetting in continual LLM fine-tuning by identifying and isolating a small, task-specific subset of essential parameters for each task.

View →
cs.CLRecentMay 29, 2026

TRACE: Discovering Task-Specific Parameter via Adaptation-Aware Probing for Continual Fine-Tuning

Xiaosong Han, Ke Chen, Xindi Dai, Di Liang +6 more

TRACE proposes a novel method to mitigate catastrophic forgetting in continual LLM fine-tuning by identifying and isolating a small, task-specific subset of essential parameters for each task.

View →
cs.CYcs.AIcs.CRMar 26, 2026

Preserving Decision Sovereignty in Military AI: A Trade-Secret-Safe Architectural Framework for Model Replaceability, Human Authority, and State Control

Peng Wei, Wesley Shu

The paper proposes the Energetic Paradigm, a model-agnostic architectural framework that allows states to maintain decision sovereignty and control over military AI systems, even when using proprietar…

View →
cs.AIcs.CRcs.CYMar 26, 2026

A Public Theory of Distillation Resistance via Constraint-Coupled Reasoning Architectures

Peng Wei, Wesley Shu

The paper proposes a theoretical framework, called constraint-coupled reasoning, to make AI models less susceptible to knowledge distillation by coupling high-level capabilities to internal stability…

View →