Skip to content

Author

Feixiang Liu

We have 2 of 12 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

SEER: A Self-Grounded Evidence Interface for Controlled Spatial Relation Classification

SEER (Self-grounded Evidence for Entity-Relation Reasoning), a training-free inference-time evidence interface for frozen VLMs, is proposed and established as the principal intervention, with reciprocal consistency as a smaller protocol-specific refinement.

Feixiang Liu, Likun Wang, Qiang Qiu et al. · 0 citations
Preprint Jul 2026

Beyond Accuracy: Auditing Spatial Provenance in Visual Token Pruning for OCR-Critical MLLM Inference

This work turns an evidence-risk audit of text-rich multimodal large language models into an evidence-risk audit that couples answer behavior with geometric token-origin provenance, interventions, and realized cost; transparent training-free selectors isolate controlled operating points.

Feixiang Liu, Qiang Qiu, Hao Zhang et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.