1 citations · 1 across the 6 of their papers we have counts for
7 papers · 1 filter
FVQKD: A Formal Verification of Quantum Key Distribution Security in Rocq
Ruixuan Deng, Zhengfeng Ji, Jiayu Zhang +1
Quantum key distribution (QKD), whose security is derived from quantum mechanics rather than computational assumptions, is moving into standardization and commercial deployment. Ye…
Transformers Stop Thinking Too Early, and a Tiny LoRA Fixes It
Zehao Jin, Ruixuan Deng, Junran Wang
Pretrained transformers use little of their depth to follow references in context. Thirteen base models reliably follow only 1.4-3.6 lines, and extra pretrained loops add little. A…
Persona Dosing: Calibrated Activation Steering for Graded Trait Control
Zehao Jin, Junran Wang, Ruixuan Deng +4
An activation-steering coefficient sets intervention strength, but requesting a particular degree of persona expression requires a behavioral scale. We study persona dosing: contro…
Long-horizon autoformalization of a core theorem underlying MIP* = RE
Sirui Lu, Ruixuan Deng, David Zhu +2
Landmark mathematical formalizations have taken specialist teams years to complete. We present FormalFlow, a system that coordinates AI proving agents under human supervision to ad…
How Benchmarks Mis-Score Computer-Use Agents
Zihan Dong, Zhiyuan Ma, Zekun Wang +5
Computer-use agents (CUA) are being deployed to browse the web and operate desktop software, yet their benchmark scores are still commonly produced by brittle scripted oracles. A s…
Decompose Sparsely Where You Should, Absorb Densely Where You Should No
Ruixuan Deng, Zehao Jin, Zekun Wang +1
Sparse autoencoders (SAEs) are typically trained to reconstruct the \textbf{entire} residual stream through a sparse dictionary, implicitly assuming that all activation content is…
Beyond Steering Vector: Flow-based Activation Steering for Inference-Time Intervention
Zehao Jin, Ruixuan Deng, Junran Wang +2
Activation steering has emerged as a promising alternative for controlling language-model behavior at inference time by modifying intermediate representations while keeping model p…