Fan Zhang

SEP 9, 2026 · PREPRINT

Maverick: Private and Verifiable LLM Inference Made Practical via Matrix-Vector Multiplication Delegation

arXiv

An early-stage prototype implementation of a novel cryptographic protocol for LLM inference, demonstrating feasibility and throughput gains but lacking peer review, real-world deployment validation, or comparison against established privacy-preserving inference methods.

AUG 10, 2026 · PREPRINT

RA-FinBERT: Rule-aware LoRA adaptation for low-resource financial sentiment classification

arXiv

Single-dataset computational study showing improved accuracy on a sentiment classification task, but lacks peer review, clinical validation, comparative rigor, and generalization evidence across datasets or domains.