Neuromorphic Computing Group

Brain-Inspired Systems at UC Santa Cruz

Neuromorphic Computing Group

Monthly Archives: February 2024

New Preprint: “Addressing cognitive bias in medical language models” led by Ph.D. Candidate Samuel Schmidgall

Posted on by

Preprint link here.

Abstract: The integration of large language models (LLMs) into the medical field has gained significant attention due to their promising accuracy in simulated clinical decision-making settings. However, clinical decision-making is more complex than simulations because physicians’ decisions are shaped by many factors, including the presence of cognitive bias. However, the degree to which LLMs are susceptible to the same cognitive biases that affect human clinicians remains unexplored. Our hypothesis posits that when LLMs are confronted with clinical questions containing cognitive biases, they will yield significantly less accurate responses compared to the same questions presented without such biases.In this study, we developed BiasMedQA, a novel benchmark for evaluating cognitive biases in LLMs applied to medical tasks. Using BiasMedQA we evaluated six LLMs, namely GPT-4, Mixtral-8x70B, GPT-3.5, PaLM-2, Llama 2 70B-chat, and the medically specialized PMC Llama 13B. We tested these models on 1,273 questions from the US Medical Licensing Exam (USMLE) Steps 1, 2, and 3, modified to replicate common clinically-relevant cognitive biases. Our analysis revealed varying effects for biases on these LLMs, with GPT-4 standing out for its resilience to bias, in contrast to Llama 2 70B-chat and PMC Llama 13B, which were disproportionately affected by cognitive bias. Our findings highlight the critical need for bias mitigation in the development of medical LLMs, pointing towards safer and more reliable applications in healthcare.

New Paper: “Surgical Gym: A high-performance GPU-based platform for reinforcement learning with surgical robots” led by PhD Candidate Samuel Schmidgall accepted at the 2024 IEEE Intl. Conf. on Robotics and Automation (ICRA 2024)

Posted on by

Preprint link here.

Abstract: Recent advances in robot-assisted surgery have resulted in progressively more precise, efficient, and minimally invasive procedures, sparking a new era of robotic surgical intervention. This enables doctors, in collaborative interaction with robots, to perform traditional or minimally invasive surgeries with improved outcomes through smaller incisions. Recent efforts are working toward making robotic surgery more autonomous which has the potential to reduce variability of surgical outcomes and reduce complication rates. Deep reinforcement learning methodologies offer scalable solutions for surgical automation, but their effectiveness relies on extensive data acquisition due to the absence of prior knowledge in successfully accomplishing tasks. Due to the intensive nature of simulated data collection, previous works have focused on making existing algorithms more efficient. In this work, we focus on making the simulator more efficient, making training data much more accessible than previously possible. We introduce Surgical Gym, an open-source high performance platform for surgical robot learning where both the physics simulation and reinforcement learning occur directly on the GPU. We demonstrate between 100-5000x faster training times compared with previous surgical learning platforms. The code is available at: https://github.com/SamuelSchmidgall/SurgicalGym.

OpenNeuromorphic Talk: “Neuromorphic Intermediate Representation”

Posted on by

In this workshop, we will show you how to move models from your favourite framework directly to neuromorphic hardware with 1-2 lines of code. We will present the technology behind, the Neuromorphic Intermediate Representation , and demonstrate how we can use it to run a live spiking convnet on the Speck chip.
Presented by Jens Pedersen, Bernhard Vogginger, Felix Bauer, and Jason Eshraghian. See the recording here.

New Paper: “Exploiting deep learning accelerators for neuromorphic workloads” led by Ph.D. Candidate Vincent Sun in collaboration with Graphcore published in Neuromorphic Computing and Engineering

Posted on by

See the full paper here.

Abstract

Spiking neural networks (SNNs) have achieved orders of magnitude improvement in terms of energy consumption and latency when performing inference with deep learning workloads. Error backpropagation is presently regarded as the most effective method for training SNNs, but in a twist of irony, when training on modern graphics processing units (GPUs) this becomes more expensive than non-spiking networks. The emergence of Graphcore’s Intelligence Processing Units (IPUs) balances the parallelized nature of deep learning workloads with the sequential, reusable, and sparsified nature of operations prevalent when training SNNs. IPUs adopt multi-instruction multi-data (MIMD) parallelism by running individual processing threads on smaller data blocks, which is a natural fit for the sequential, non-vectorized steps required to solve spiking neuron dynamical state equations. We present an IPU-optimized release of our custom SNN Python package, snnTorch, which exploits fine-grained parallelism by utilizing low-level, pre-compiled custom operations to accelerate irregular and sparse data access patterns that are characteristic of training SNN workloads. We provide a rigorous performance assessment across a suite of commonly used spiking neuron models, and propose methods to further reduce training run-time via half-precision training. By amortizing the cost of sequential processing into vectorizable population codes, we ultimately demonstrate the potential for integrating domain-specific accelerators with the next generation of neural networks.

New snnTorch Tutorial: Exoplanet Hunter by Undergraduate Students Ruhai Lin, Aled dela Cruz, and Karina Aguilar

Posted on by

See the tutorial here.

The transit method is a widely used and successful technique for detecting exoplanets. When an exoplanet transits its host star, it causes a temporary reduction in the star’s light flux (brightness). Compared to other techniques, the transmit method has has discovered the largest number of planets.

Astronomers use telescopes equipped with photometers or spectrophotometers to continuously monitor the brightness of a star over time. Repeated observations of multiple transits allows astronomers to gather more detailed information about the exoplanet, such as its atmosphere and the presence of moons.

Space telescopes like NASA’s Kepler and TESS (Transiting Exoplanet Survey Satellite) have been instrumental in discovering thousands of exoplanets using the transit method. Without the Earth’s atmosphere in the way, there is less interference and more precise measurements are possible. The transit method continues to be a key tool in advancing our understanding of exoplanetary systems. For more information about transit method, you can visit NASA Exoplanet Exploration Page.