Institution profile

University of Texas Southwestern Medical Center

Academic institutionnorthamerica · us
Official website
Research library94linked papers
Opportunities0open roles
Selected work

Representative Papers

Scaling Medical Reasoning Verification via Tool-Integrated Reinforcement Learning

Jan 28, 2026

This work addresses the limitations of existing medical reasoning verification methods, which rely on single-step retrieval and provide only scalar rewards, thereby lacking interpretability and dynamic knowledge acquisition. The authors propose a tool-augmented reinforcement learning agent framework that iteratively queries external medical corpora during verification and integrates trajectory-supervised iterative reinforcement learning with an adaptive curriculum mechanism. This approach enables dynamic evidence retrieval for the first time, substantially improving both verification efficiency and reliability. Evaluated on four medical reasoning benchmarks, the method significantly outperforms current state-of-the-art approaches, achieving a 23.5% absolute accuracy gain on MedQA and a 32.0% improvement on MedXpertQA, while reducing the required sampling budget by a factor of eight.

1 citationsRead paper

LLM-as-RNN: A Recurrent Language Model for Memory Updates and Sequence Prediction

Jan 19, 2026

This work addresses the limitation of large language models (LLMs) in lacking updatable memory during inference, which hinders their ability to dynamically correct errors. The authors propose the LLM-as-RNN framework, which— for the first time—recursively employs a frozen LLM by constructing a natural-language memory state through structured system prompts and iteratively rewriting this memory based on feedback at each step, thereby enabling online learning without any parameter updates. This approach simultaneously enhances interpretability and performance, achieving an average accuracy improvement of 6.5% across three sequential prediction tasks in healthcare, meteorology, and finance. It significantly outperforms baseline methods including zero-shot prompting, full-history prompting, and MemPrompt, while also producing human-readable learning trajectories that elucidate the model’s reasoning process.

1 citationsRead paper
Recent publications

Latest Papers

Extended KAFR: A kinematic-adaptive paradigm for the efficient analysis of surgical video

Aug 02, 2026

This work addresses the computational burden in laparoscopic surgery video analysis caused by redundant duration, motion artifacts, and low visual quality. To tackle this, the authors propose an adaptive keyframe selection method driven by surgical instrument motion dynamics. Specifically, a fine-tuned YOLO model detects and segments instruments, and keyframes are dynamically selected based on their displacement or velocity changes. These frames are then fed into an X3D model for surgical phase classification. Evaluated on the Cholec80 dataset, the method achieves a 91.0% F1 score using only 0.58% of the total frames—approximately seven times fewer than the conventional 4% sampling rate—matching the performance of state-of-the-art models such as LoViT and Trans-SVNet. This study marks the first successful application of kinematics-driven adaptive sampling to complex laparoscopic scenarios, enabling highly accurate phase recognition at extremely low sampling rates.

0 citationsRead paper