26 papers
Diagnosing Under-Development of Irreversible Processes in Video Generation
Jian Xu, Yanning Wu, Delu Zeng +2
Many physical attributes are \emph{irreversible}: ice melts but does not re-freeze, paper chars but does not un-burn. Do video generators respect this? We show the question is hard…
Look Again Before You Abstain:Budgeted Conformal Evidence Acquisition for Reliable Vision-Language Model
Jian Xu, Delu Zeng, Yanning Wu +2
Large vision-language models (LVLMs) hallucinate: they assert visual details that the image does not support. A principled remedy is selective prediction with a distribution-free g…
Fixed-Protocol Amortized MPS Tomography with Conformalized Predictive Uncertainty
Jian Xu, Delu Zeng, John Paisley +1
The paper introduces a fixed‑protocol amortized estimator for matrix‑product‑state quantum tomography that leverages an informative local Pauli measurement design and a gauge‑invar…
Entanglement as a Structural Complexity Axis: A PAC-Bayesian View of Generalization in Quantum Policies and Value Functions
Jian Xu, Delu Zeng, John Paisley +1
Parameterized quantum circuits (PQCs) are increasingly used as policies and value functions in quantum reinforcement learning, yet it remains unclear when and why quantum policies…
When Can You Debias an LLM Judge? Identifiability Limits, a Test, and Designs for Top-k Ranking
Jian Xu, Delu Zeng, John Paisley +1
Large language models (LLMs) are increasingly used as cheap, scalable judges that compare candidate outputs pairwise. Because such judges prefer verbose or well-formatted answers,…
Calibration, Not Compilation: Detecting and Repairing Misspecified Probabilistic Programs Written by Language Models
Jian Xu, Delu Zeng, John Paisley +1
Language models increasingly write probabilistic programs (in NumPyro, Stan, or Pyro), but a program that compiles, runs, and passes every unit test can still be \emph{statisticall…