TecoPrompt: Temporal-Conservative Prompt Learning for Vision-Language Models

2026-09-15 · Zeyi Shao, Haowen Hua, Jiaxin Zhang, John See, Zeyd Boukhers, Cong Yang arxiv

Prompt learning adapts vision-language models, such as CLIP, by adjusting a small set of context tokens. However, under few-shot supervision, even moderate label noise can disrupt prompt optimization. To address this issue, we propose TecoPrompt, a closed-loop robust prompt-learning framework that revisits optimal transport (OT) pseudo-labeling from a temporal perspective. TecoPrompt employs an entropic OT plan in the CLIP semantic space to obtain globally consistent label candidates. It verifies the reliability of these candidates by examining trajectory stability: a noisy label is only rewritten if the OT candidate remains unchanged within a K-epoch temporal stability window and passes a confidence gate based on Exponential Moving Average (EMA). This approach helps reduce confirmation bias. The rewritten labels are then integrated back into prompt training using a tri-group objective that includes three loss functions aligned with clean, mid, and noisy subsets. Experiments on seven datasets with synthetic symmetric and asymmetric noise, as well as Food101N, demonstrate significant performance improvements. For example, on the OxfordPets dataset, with 50% asymmetric noise, TecoPrompt achieves an accuracy of 0.843, up from 0.775.

📄 PDF Abstract BibTeX arXiv:2609.16858

Code (0)

등록된 구현이 없습니다.

Conservative Bias in Large Language Models: Measuring Relation Predictions

2025-06-09 · Toyin Aguda, Erik Wilson, Allan Anzagira, Simerjot Kaur 외

Large language models (LLMs) exhibit pronounced conservative bias in relation extraction tasks, frequently defaulting to No_Relation label when an appropriate option is unavailable. While this behavior helps prevent inco…

Hallucination Relation Relation Extraction Semantic Similarity+1

Are Multimodal LLMs Ready for Surveillance? A Reality Check on Zero-Shot Anomaly Detection in the Wild

2026-03-05 · Shanle Yao, Armin Danesh Pazho, Narges Rashvand, Hamed Tabkhi arxiv

Multimodal large language models (MLLMs) have demonstrated impressive general competence in video understanding, yet their reliability for real-world Video Anomaly Detection (VAD) remains largely unexplored. Unlike conve…

Video Anomaly Detection Binary Classification

Minimal-Edit Instruction Tuning for Low-Resource Indic GEC

2025-11-28 · Akhil Rajeev P arxiv

Grammatical error correction for Indic languages faces limited supervision, diverse scripts, and rich morphology. We propose an augmentation-free setup that uses instruction-tuned large language models and conservative d…

parameter-efficient fine-tuning Grammatical Error Correction

Risk-aware Selective Prompting for Hallucination Mitigation in Large Vision-Language Models

2026-05-27 · Yuang Huang, Yafeng Zhang, Yu Zilan arxiv

Prompt-based verification is widely used to mitigate hallucinations in large vision-language models (LVLMs), yet when it helps remains poorly understood. We systematically study verification prompting across two represen…

Visual Grounding

Multi-Turn Adaptive Prompting Attack on Large Vision-Language Models

2026-02-16 · In Chong Choi, Jiacheng Zhang, Feng Liu, Yiliao Song arxiv

Multi-turn jailbreak attacks have proven effective against text-only large language models (LLMs), where malicious content is gradually introduced to bypass safety alignment. However, effectively extending such attacks t…