Search for: All records

While large language models (LLMs) have demonstrated impressive capabilities across tasks in language understanding and interactive decision making, their abilities for reasoning (e.g. chain-of-thought prompting) and acting (e.g. action plan generation) have primarily been studied as separate topics. In this paper, we explore the use of LLMs to generate both reasoning traces and task-specific actions in an interleaved manner, allowing for greater synergy between the two: reasoning traces help the model induce, track, and update action plans as well as handle exceptions, while actions allow it to interface with external sources, such as knowledge bases or environments, to gather additional information. We apply our approach, named ReAct, to a diverse set of language and decision making tasks and demonstrate its effectiveness over state-of-the-art baselines, as well as improved human interpretability and trustworthiness over methods without reasoning or acting components. Concretely, on question answering (HotpotQA) and fact verification (Fever), ReAct overcomes issues of hallucination and error propagation prevalent in chain-of-thought reasoning by interacting with a simple Wikipedia API, and generates human-like task-solving trajectories that are more interpretable than baselines without reasoning traces. On two interactive decision making benchmarks (ALFWorld and WebShop), ReAct outperforms imitation and reinforcement learning methods by an absolute success rate of 34% and 10% respectively, while being prompted with only one or two in-context examples. 

Phototherapy represents an attractive route for treating a range of challenging dermatological diseases. Existing skin phototherapy modalities rely on direct UV illumination, although with limited efficacy in addressing disorders of deeper tissue and with requirements for specialized illumination equipment and masks to shield unaffected regions of the skin. This work introduces a skin‐integrated optoelectronic device that incorporates an array of UVA (360 nm) light emitting diodes in layouts that match those of typical lesional plaques and in designs that couple to biocompatible, penetrating polymer microneedle light waveguides to provide optical access to deep skin. Monte Carlo simulations and experimental results in phantom skin suggest that these waveguides significantly enhance light delivery to deep skin, with a >4‐fold increase for depths of >500 µm. In ex vivo human skin, the devices show reduced measures of phototoxicity compared to direct illumination and enhanced modulation of gene expression relevant to sclerosing skin diseases. These systems are also compatible with design principles in soft, skin‐compatible electronics and battery‐powered wireless operation. Collectively, the favorable mechanical and light delivery properties of these devices expand possibilities in targeting of deep skin lesions beyond those attainable with clinical‐standard UV light therapy approaches.