content type videopublished February 23, 2026
Apple Workshop on Reasoning and Planning 2025: On Evaluating and Improving Robustness in Visual Reasoning Abilities of Vision-Language Models
AuthorsMelanie Mitchell (Santa Fe Institute)
Related readings and updates.
LEAD: Breaking the No-Recovery Bottleneck in Long-Horizon Reasoning
July 24, 2026research area Methods and Algorithms, research area Speech and Natural Language Processing
Long-horizon execution in Large Language Models (LLMs) remains unstable even when high-level strategies are provided. Evaluating on controlled algorithmic puzzles, we demonstrate that while decomposition is essential for stability, extreme decomposition creates a “no-recovery bottleneck”. We show that this bottleneck becomes critical due to highly non-uniform error distribution, where consistent errors on a few “hard” steps become irreversible…
Environment-free Synthetic Data Generation for API-Calling Agents
July 21, 2026research area Methods and Algorithms
Training API-calling large language model (LLM) agents demands massive amounts of high-quality trajectories. However, collecting such data at scale typically requires fully implemented environments with executable APIs and realistic, pre-populated backend databases, creating a major bottleneck for scalability. To overcome this, we propose an environment-free synthetic data generation approach that leverages LLMs as on-the-fly digital world…
Discover opportunities in Machine Learning.
Our research in machine learning breaks new ground every day.