THEMIS: Towards Holistic Evaluation of MLLMs for Scientific Paper Fraud Forensics

TL;DR

THEMIS is a comprehensive benchmark designed to evaluate multimodal large language models on complex visual fraud reasoning in real-world academic scenarios, highlighting current model limitations.

Contribution

The paper introduces THEMIS, a novel multi-task benchmark with diverse scenarios, fraud types, and fine-grained manipulations for rigorous evaluation of MLLMs.

Findings

Even the best model, GPT-5, scores only 56.15% overall.

The benchmark covers over 4,000 questions from real retracted papers.

Models struggle with complex textures and multiple manipulations.

Abstract

We present THEMIS, a novel multi-task benchmark designed to comprehensively evaluate multimodal large language models (MLLMs) on visual fraud reasoning within real-world academic scenarios. Compared to existing benchmarks, THEMIS introduces three major advances. (1) Real-World Scenarios and Complexity: Our benchmark comprises over 4,000 questions spanning seven scenarios, derived from authentic retracted-paper cases and carefully curated multimodal synthetic data. With 60.47% complex-texture images, THEMIS bridges the critical gap between existing benchmarks and the complexity of real-world academic fraud. (2) Fraud-Type Diversity and Granularity: THEMIS systematically covers five challenging fraud types and introduces 16 fine-grained manipulation operations. On average, each sample undergoes multiple stacked manipulation operations, with the diversity and difficulty of these manipulations demanding a high level of visual fraud reasoning from the models. (3) Multi-Dimensional Capability Evaluation: We establish a mapping from fraud types to five core visual fraud reasoning capabilities, thereby enabling an evaluation that reveals the distinct strengths and specific weaknesses of different models across these core capabilities. Experiments on 16 leading MLLMs show that even the best-performing model, GPT-5, achieves an overall performance of only 56.15%, demonstrating that our benchmark presents a stringent test. We expect THEMIS to advance the development of MLLMs for complex, real-world fraud reasoning tasks.