# Apple — Selection-Based Structured Reasoning: Toward Efficient Multimodal Search Agents

- Company: Apple (apple.com)
- Announced: 2026-10-15
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://machinelearning.apple.com/research/selection-structured-reasoning
- Record: https://forck.live/items/23073-selection-based-structured-reasoning-toward-efficient-multimodal-search-agents
- Subject: Machine Learning Research

Apple researchers introduced Selection-based Structured Reasoning (SSR), a framework that reformulates reasoning as selection from pre-specified natural-language candidates rather than open-ended generation. Evaluated on seven multimodal search benchmarks using 2B and 4B models, SSR achieved success rates competitive with leading search agents of the same scale while reducing per-turn reasoning latency by over 90% and total per-question model inference latency by 28–54%.

## Evidence

Verbatim from https://machinelearning.apple.com/research/selection-structured-reasoning:

> SSR achieves an average success rate competitive with leading search agents of the same scale, while reducing per-turn reasoning latency by over 90% and total per-question model inference latency by 28–54%.

---

Record: https://forck.live/items/23073-selection-based-structured-reasoning-toward-efficient-multimodal-search-agents
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md