# A Deterministic Evidence Layer for Vision-Language Autism Screening from Naturalistic Home Video

arXiv: https://arxiv.org/abs/2610.09217 · HTML: https://mlsanity.com/paper/2610.09217/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 66 references that are in the map.
- First posted on arXiv 6 Oct 2026.

## Builds on (most cited first, 9 of 9)

- [τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains](https://mlsanity.com/paper/2406.12045/) (2024)
- [Reasoning-Aligned Perception Decoupling for Scalable Multi-modal Reasoning](https://mlsanity.com/paper/2506.04559/) (2025)
- [VideoZeroBench: Probing the Limits of Video MLLMs with Spatio-Temporal Evidence Verification](https://mlsanity.com/paper/2604.01569/) (2026)
- [Video-based Analysis Reveals Atypical Social Gaze in People with Autism Spectrum Disorder](https://mlsanity.com/paper/2409.00664/) (2024)
- [HERO: Hierarchical Evidential Reasoning Optimization for Radiology Report Generation via Reason-then-Summarize](https://mlsanity.com/paper/2601.03321/) (2026)
- [Towards Child-Inclusive Clinical Video Understanding for Autism Spectrum Disorder](https://mlsanity.com/paper/2409.13606/) (2024)
- [Dep-LLM: Training-Free Depression Diagnosis via Evidence-Guided Structured Multi-factor with Reliable LLM Reasoning](https://mlsanity.com/paper/2606.10796/) (2026)
- [DiagnoLLM: A Hybrid Bayesian Neural Language Framework for Interpretable Disease Diagnosis](https://mlsanity.com/paper/2511.05810/) (2025)
- [Fine-tuning a multimodal large language model for clinician-grade autism behavioral scoring from short home videos](https://mlsanity.com/paper/2606.27484/) (2026)

## Abstract (arXiv)

Autism spectrum disorder (ASD) is diagnosed through specialist observation of a child's social behavior, and access to that expertise is the bottleneck for early identification. Vision-language models (VLMs) describe a child's behavior from video well; the verdict drawn from the description is unstable: at temperature~0, across eight pipeline configurations on one backbone, 16--37% of clips change their predicted label between repeated runs, and the cause lies in the serving stack. We keep the VLM frozen and move the decision out of the model. Under our grounded perception constraints the VLM writes an event table of timestamped, glossary-labeled events that names the eliciting press and logs counter-evidence; a text-only stage age-calibrates the confidence of each row; a deterministic weight-of-evidence scorer sums it into an evidence total and stratifies it into a risk category, so every decision decomposes into named per-feature contributions and can be re-scored from the saved table. On 43 caregiver-recorded, protocol-free home free-play clips of preschool children, the pipeline reaches AUC 0.851 ± 0.012, 86.0% accuracy, and F1 71.8 over three runs. It labels 74.4% of clips correctly in every run (60.5% for the zero-shot baseline) and flags no typically developing clip in every run (9 of 31 at zero-shot). An ablation on the same backbone attributes the gain to the grounded perception constraints read through the deterministic scorer.