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

Cheng Tan

3 indexed papers

Recent (6 mo)
2
With code
0
Influential cites
0
Benchmarked
0

Publications per year

3
26

Top categories

AI×2OS×1Vision×1Crypto×1

Frequent co-authors

Bo Yin1×
Xiaobin Hu1×
Chengming Xu1×
Ruolin Shen1×
Mo Yang1×
Jiangning Zhang1×

Research Timeline

2026
Toward a Principled Framework for Agent Safety Measurement

The paper introduces BOA, a novel framework that measures agent safety by exhaustively searching the entire in-budget trajectory space, thereby identifying unsafe behaviors missed by traditional sampling methods.

SPOT-E: Test-Time Entropy Shaping with Visual Spotlights for Frozen VLMs

This paper introduces SPOT-E, a plug-and-play test-time method that improves the performance of vision-language models on evidence-intensive tasks by producing question-conditioned spotlights.

Defining AI-Native Systems: Autonomy as Revision Authority

This paper proposes a definition for 'AI-nativeness' in systems, based on an AI agent's authority over system decisions.

Highlighted terms show continued research focus across papers

Papers

cs.AIcs.OSTheoreticalRecentJul 22, 2026

Defining AI-Native Systems: Autonomy as Revision Authority

Cheng Tan

This paper proposes a definition for 'AI-nativeness' in systems, based on an AI agent's authority over system decisions.

View →
cs.CVcs.AIEmpiricalRecent
Jun 18, 2026

SPOT-E: Test-Time Entropy Shaping with Visual Spotlights for Frozen VLMs

Bo Yin, Xiaobin Hu, Chengming Xu, Ruolin Shen +5 more

This paper introduces SPOT-E, a plug-and-play test-time method that improves the performance of vision-language models on evidence-intensive tasks by producing question-conditioned spotlights.

View →
cs.CRMay 2, 2026

Toward a Principled Framework for Agent Safety Measurement

Shuyi Lin, Anshuman Suri, Alina Oprea, Cheng Tan

The paper introduces BOA, a novel framework that measures agent safety by exhaustively searching the entire in-budget trajectory space, thereby identifying unsafe behaviors missed by traditional sampl…

View →