Author

Zhangheng Li

h-index 6 View on Semantic Scholar

We have 2 of 15 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Aug 2026

HPFA: Hypergraph-Based Paired Failure Attribution for LLM Reasoning

A hypergraph-based paired failure attribution (HPFA) framework that attributes the failure root cause by comparing the hyperedges of the targeted failure reasoning path against a reference successful path is proposed, and the trained attributor consistently improves reasoning accuracy at test time.

Runchuan Zhu, Hong Dung Lai, Bowen Jiang et al. · 1 citation

Preprint Jul 2026

POPS: Recovering Unlearned Multi-Modality Knowledge in MLLMs with Prompt-Optimized Parameter Shaking

This paper proposes a novel adversarial strategy, namely Prompt-Optimized Parameter Shaking (POPS), aiming to recover the supposedly unlearned multi-modality knowledge from the MLLMs, exposing fundamental vulnerabilities that challenge the foundational robustness of representative MMU-based privacy protections.

Zhangheng Li, Jianing Zhu, Junyuan Hong et al. · 0 citations