Peishun Liu
3 papers · 2026–2026 · 1 venue · across top CS/AI venues
Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.
Venues
Research topics
Keywords
vision-language model
(3)
gender bia
(2)
visual grounding
(1)
cross attention
(1)
backdoor adjustment
(1)
large language model
(1)
causal mediation analysis
(1)
hallucinations
(1)
contrastive decoding probes
(1)
decoding-time penalties
(1)
token-level contrastive grounding score
(1)
activation replacement
(1)
attention mechanism
(1)
layer-by-layer
(1)
token selection
(1)
Papers
Mechanism-Level Evaluation for Vision-Language Models: Controlled Activation-Replacement Diagnosis of Gender Bias
EMNLP 2026
ViD: Vision-Dominant Gender Bias Mitigation for Large Vision-Language Models
EMNLP 2026
What Do Hallucinations Reveal About Multimodal Reasoning? Diagnosing Visual Grounding Failures via Contrastive Decoding Probes
EMNLP 2026