Explanation-Bound Tool Execution (EBTE) is a mediation layer that converts unstructured rationales from AI agents into typed action claims to ensure server-held intent and policy compliance. EBTE is tested in 136 conformance scenarios and a frozen historical record with promising results.

Researchers introduced SpecOpt, a molecular design task to improve the specificity of existing compounds by suggesting structural modifications that increase binding preference for an intended target over off-targets. An agentic framework using LLMs was developed to propose targeted modifications. The method achieved significant improvements in target-off-target binding gap for 84.8% of compounds, establishing a new molecular design problem.

Researchers propose Test-Time Self-Evolving via Reflection (TTSR), a self-evolving framework for adapting large language models during inference using unlabeled test inputs. TTSR improves upon existing methods by addressing bottlenecks in learnable samples and efficient exploration. The framework alternates between a Student and Teacher role, with the Teacher analyzing failed trajectories and synthesizing targeted questions. Experiments show consistent test-time improvements and strong cross-backbone generalization.

A new specification (PACE) for AI cinematic expression aims to bridge the gap between screenplay and film planning by providing a typed representation of the plan, including character and prop locations, camera positions, and other key details. This can improve the accuracy of AI-generated images and videos.

This research paper proposes a unified framework for Riemannian deep learning, generalizing batch normalization and multinomial logistic regression to broad classes of Lie groups and gyrogroups. It also introduces novel neural networks for geometric representations, including hyperbolic space and full-rank correlation matrices. This work has potential implications for developing more robust and adaptive AI models.