NemotronLabs VoiceChat is an open full-duplex speech-to-speech model that integrates speech recognition, generation, and external tool use in a single model. It achieves state-of-the-art results on various benchmarks, demonstrating the potential for real-time conversational behavior. This development is relevant to AI agent builders, as it showcases advancements in speech-to-speech models that can be applied to agent capabilities.

Researchers propose a reinforcement learning framework, PA-RL, that uses artificial potential fields as an action representation for contact-rich robotic manipulation tasks. This approach allows the policy to adapt task strategy and low-level motion generation, reducing learning burden and improving performance. PA-RL achieves 100% evaluation success rate in simulation and demonstrates real-robot deployment feasibility.

A new routing framework, Low-rank and Lipschitz-controlled Routing (L2R), is proposed for Mixture-of-Experts (MoE) models. L2R reshapes the routing space and scoring geometry, improving routing discriminability and stable expert specialization. This matters to people building AI agents as it could lead to improved model performance and more efficient expert expressiveness.

Researchers propose FAMOS, a feed-forward model that predicts movable-part segmentation and joint parameters from sparse, unordered point clouds. The model jointly reasons over multiple observations and introduces a Multi-state Articulation Transformer for aggregating articulation cues. The authors also introduce a procedural data generator for synthesizing self-annotated assets during training.

Jev, a probabilistic decision model, has reached record-breaking adoption on AI Gateway, demonstrating its speed and efficiency in production. It was adopted by nearly 13% of paid teams within 24 hours, outperforming previous models. Jev's unique architecture enables it to provide typed answers directly usable by code, making it a promising solution for developers to make structured decisions in software.

A new hierarchical long-horizon vision-language-action architecture with an explicit language-memory module is proposed to improve the success rate and robustness of VLA models on complex tasks. The architecture decouples the system into a high-level VLM and a low-level VLA, enabling persistent temporal tracking and dynamic correction during long-horizon execution.

A research paper proposes a two-pass decomposition approach for multimodal review tasks, where the first pass transcribes the source and the second pass reviews the transcript. This approach improves faithfulness and coverage, but introduces potential failure modes such as running out of room or confabulation. This matters to people building or operating AI agents because it suggests a potential solution to the limitations of single-pass multimodal models.

Researchers propose a lightweight CNN-integrated Compact Convolutional Transformer (CCT) model for breast cancer mammography image detection and classification, achieving high accuracy with a low number of parameters.

Researchers proposed R-U-Net, an ECG delineation model that pairs ResNet-18 with a U-Net decoder, achieving improved performance over previous models. This advancement has implications for AI models interpreting ECGs and may inform the design of AI architectures for similar tasks.

Researchers propose a new AI model, TDGP, for Audio-Visual Navigation (AVN) that uses Transformer-based token fusion and dynamic graph planning to improve agent navigation in environments with incomplete or misleading visual perception.

Researchers proposed a framework to generate context-specific large language model (LLM) benchmark datasets by combining expert input with synthetic data generation. This approach aims to balance validity and scalability, addressing the limitations of existing benchmark construction methods. The framework uses a schema to elicit key information about evaluation tasks and guides synthetic data generation with four measurement validity criteria: coverage, diversity, content realism, and stylistic realism.