English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Vision Language Models (VLMs) have demonstrated strong reasoning capabilitiesin Visual Question Answering (VQA) tasks; however, their ability to performTheory of Mind (ToM) tasks, such as inferring human intentions, beliefs, andmental states, remains underexplored. We propose an open-ended questionframework to evaluate VLMs' performance across diverse categories of ToM tasks.We curated and annotated a benchmark dataset of 30 images and evaluated theperformance of four VLMs of varying sizes. Our results show that the GPT-4model outperformed all the others, with only one smaller model, GPT-4o-mini,achieving comparable performance. We observed that VLMs often struggle to inferintentions in complex scenarios such as bullying or cheating. Our findingsreveal that smaller models can sometimes infer correct intentions despiterelying on incorrect visual cues. The dataset is available athttps://github.com/ximingwen/ToM-AAAI25-Multimodal.

How Well Can Vison-Language Models Understand Humans' Intention? An Open-ended Theory of Mind Question Evaluation Benchmark