# Can Motion-Language Models Ground Structure? STRIDE for Evaluating the Evaluators

arXiv: https://arxiv.org/abs/2609.32462 · HTML: https://mlsanity.com/paper/2609.32462/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 34 of its 42 references that are in the map.
- First posted on arXiv 26 Sep 2026.

## Builds on (most cited first, 30 of 34)

- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [Human Motion Diffusion Model](https://mlsanity.com/paper/2209.14916/) (2022)
- [MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model](https://mlsanity.com/paper/2208.15001/) (2022)
- [When and why vision-language models behave like bags-of-words, and what to do about it?](https://mlsanity.com/paper/2210.01936/) (2022)
- [T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations](https://mlsanity.com/paper/2301.06052/) (2023)
- [Executing your Commands via Motion Diffusion in Latent Space](https://mlsanity.com/paper/2212.04048/) (2022)
- [Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality](https://mlsanity.com/paper/2204.03162/) (2022)
- [MotionGPT: Human Motion as a Foreign Language](https://mlsanity.com/paper/2306.14795/) (2023)
- [TEMOS: Generating diverse human motions from textual descriptions](https://mlsanity.com/paper/2204.14109/) (2022)
- [MotionCLIP: Exposing Human Motion Generation to CLIP Space](https://mlsanity.com/paper/2203.08063/) (2022)
- [MoMask: Generative Masked Modeling of 3D Human Motions](https://mlsanity.com/paper/2312.00063/) (2023)
- [TempCompass: Do Video LLMs Really Understand Videos?](https://mlsanity.com/paper/2403.00476/) (2024)
- [TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts](https://mlsanity.com/paper/2207.01696/) (2022)
- [ReMoDiffuse: Retrieval-Augmented Motion Diffusion Model](https://mlsanity.com/paper/2304.01116/) (2023)
- [What's "up" with vision-language models? Investigating their struggle with spatial reasoning](https://mlsanity.com/paper/2310.19785/) (2023)
- [SugarCrepe: Fixing Hackable Benchmarks for Vision-Language Compositionality](https://mlsanity.com/paper/2306.14610/) (2023)
- [TMR: Text-to-Motion Retrieval Using Contrastive 3D Human Motion Synthesis](https://mlsanity.com/paper/2305.00976/) (2023)
- [CREPE: Can Vision-Language Foundation Models Reason Compositionally?](https://mlsanity.com/paper/2212.07796/) (2022)
- [VL-CheckList: Evaluating Pre-trained Vision-Language Models with Objects, Attributes and Relations](https://mlsanity.com/paper/2207.00221/) (2022)
- [MotionLLM: Understanding Human Behaviors from Human Motions and Videos](https://mlsanity.com/paper/2405.20340/) (2024)
- [VITATECS: A Diagnostic Dataset for Temporal Concept Understanding of Video-Language Models](https://mlsanity.com/paper/2311.17404/) (2023)
- [COLA: A Benchmark for Compositional Text-to-image Retrieval](https://mlsanity.com/paper/2305.03689/) (2023)
- [Test of Time: Instilling Video-Language Models with a Sense of Time](https://mlsanity.com/paper/2301.02074/) (2023)
- [SnapMoGen: Human Motion Generation from Expressive Texts](https://mlsanity.com/paper/2507.09122/) (2025)
- [SUGARCREPE++ Dataset: Vision-Language Model Sensitivity to Semantic and Lexical Alterations](https://mlsanity.com/paper/2406.11171/) (2024)
- [MotionGPT3: Human Motion as a Second Modality](https://mlsanity.com/paper/2506.24086/) (2025)
- [MG-MotionLLM: A Unified Framework for Motion Comprehension and Generation across Multiple Granularities](https://mlsanity.com/paper/2504.02478/) (2025)
- [What is the Best Automated Metric for Text to Motion Generation?](https://mlsanity.com/paper/2309.10248/) (2023)
- [A Cross-Dataset Study for Text-based 3D Human Motion Retrieval](https://mlsanity.com/paper/2405.16909/) (2024)

## Abstract (arXiv)

Motion-language models are typically scored by motion-language evaluators, but how well these evaluators ground language structure remains unclear. Here, we introduce the Structure grounding via Temporal-order, Reflection, and Identity Diagnostic Evaluation (STRIDE) benchmark to systematically evaluate the ability of evaluators to track temporal order, mirror reflection, and action identity. STRIDE comprises 5,869 triples, each consisting of a motion, its original caption, and a perturbed caption, spanning both short and long descriptions. We likelihood-balance caption pairs to reduce text-only bias and estimate each evaluator's caption preference under unrelated motions to measure the discrimination gain from matched motions relative to this baseline. Our experiments reveal weak structural grounding and severe deficits in mirror sensitivity among the audited evaluators, which commonly used evaluation protocols fail to expose. To understand why these limitations go undetected in standard tests, we examine the evaluators more closely. We find that text-only priors alone can solve naive perturbation tests on existing datasets, while common retrieval and distributional metrics barely respond to structural corruption introduced by mirroring ground-truth motions. These findings suggest a natural intervention: structural hard negatives. Our experiments show that a simple modification to contrastive learning substantially improves performance on temporal order and mirror reflection. The benchmark and code will be released.