Vision Large Language Models (VLLMs) show promise for multi-modal understanding; yet, their application to emotion recognition remains limited by insufficient spatial and contextual awareness and by supervised fine-tuning that uniformly penalizes all reasoning tokens—suppressing valid alternatives. We present Set-of-Vision-Text Rewards (SoVTR), a unified, reward-based framework that guides VLLMs with interpretable and verifiable visual attributes, focusing optimization on the emotion-discriminative reasoning steps that truly determine correctness. SoVTR decomposes supervision into three rewards: (i) Face Reward capturing action units, gaze, and micro-expressions; (ii) Body Reward encoding pose and gesture cues; and (iii) Context Reward modeling person–object relations and scene activities. By combining BERTScore attribute matching with GRPO policy optimization, SoVTR produces attribute-grounded explanation traces while preserving non-critical reasoning diversity. Across DFEW, FERV39k, and ExpW, SoVTR consistently improves both recognition accuracy and grounding quality, delivering substantial gains over strong VLLM baselines.

Recent MLLMs have shown promise in scene-text-aware video understanding, yet their ability to understand scene text from the wearer’s perspective in AI-glasses videos remains underexplored. AI-glasses videos are passively captured without a camera-frame preview, limiting the wearer’s ability to adjust framing during recording. As a result, relevant text may be unstable, and the recorded view may not fully reflect the wearer’s attention. The challenge is thus not merely to read visible text, but to identify and read the right text given the wearer’s position, viewpoint trajectory, and goal. To systematically evaluate this capability, we introduce **WearerText**, a benchmark for **Wearer**-Centered Scene **Text** Understanding, comprising 1,101 real-world AI-glasses videos and 15,336 QA pairs spanning 13 tasks. These tasks are organized into a three-level diagnostic hierarchy: *L1 Text Perception*, *L2 Contextual Understanding*, and *L3 Wearer-Grounded Reasoning*, targeting unstable text perception, wearer-view alignment, and goal-oriented text reasoning, respectively. To ensure reliable construction and evaluation, we develop **MAH-V**, a multi-agent hybrid verification pipeline for QA construction, and propose **VerEval**, a Verified-evidence metric for open-ended Evaluation. Benchmarking 20 recent MLLMs shows that current models still struggle with wearer-centered scene text understanding, particularly on tasks requiring wearer-view alignment and goal-oriented text reasoning.

Vision Large Language Models (VLLMs) are transforming the intersection of computer vision and natural language processing; however, the potential of using visual prompts for emotion recognition in these models remains largely unexplored and untapped. Traditional methods in VLLMs struggle with spatial localization and often discard valuable global context. We propose a novel Set-of-Vision prompting (SoV) approach that enhances zero-shot emotion recognition by using spatial information, such as bounding boxes and facial landmarks, to mark targets precisely. SoV improves accuracy in face count and emotion categorization while preserving the enriched image context. Through comprehensive experimentation and analysis of recent commercial or open-source VLLMs, we evaluate the SoV model’s ability to comprehend facial expressions in natural environments. Our findings demonstrate the effectiveness of integrating spatial visual prompts into VLLMs for improving emotion recognition performance.

Mitigating bias in language models (LMs) has become a critical problem due to the widespread deployment of LMs in the industry and customer-facing applications. Numerous approaches revolve around data pre-processing and subsequent fine-tuning of language models, tasks that can be both time-consuming and computationally demanding. As alternatives, machine unlearning techniques are being explored, yet there is a notable lack of comparative studies evaluating the effectiveness of these methods. In this work, we explore the effectiveness of two machine unlearning methods: Partitioned Contrastive Gradient Unlearning (PCGU) applied on decoder models, and Negation via Task Vector, and compare them with Direct Preference Optimization (DPO) to reduce social biases in open-source LMs such as LLaMA-2 and OPT. We also implement distributed PCGU for large models. It is empirically shown, through quantitative and qualitative analyses, that negation via Task Vector method outperforms PCGU and is comparable to DPO in debiasing models with minimum deterioration in model performance and perplexity. Negation via Task Vector reduces the bias score by 25.5% for LLaMA-2 and achieves bias reduction of up to 40% for OPT models. Moreover, it can be easily tuned to balance the trade-off between bias reduction and generation quality, unlike DPO.