Usual model-agnostic explanation methods create artificial instances by slightly perturbing real instances, capturing shifts in model decisions. However, such methods rely on initial data and only provide explanations of the decision for these. To tackle these problems, we propose Therapy, the first global and model-agnostic explanation method adapted to text which requires no input dataset. Therapy generates texts following the distribution learned by a classifier through cooperative generation. Because it does not rely on initial samples, it allows to generate explanations even when data is absent (e.g., for confidentiality reasons). Moreover, conversely to existing methods that combine multiple local explanations into a global one, Therapy offers a global overview of the model behavior on the input space. Our experiments show that although using no input data to generate samples, Therapy provides insightful information about features used by the classifier that is competitive with the ones from methods relying on input samples and outperforms them when input samples are not specific to the studied model.

"Honey, Tell Me What's Wrong'', Global Explanation of Textual Discriminative Models through Cooperative Generation

Large language models (LLMs) that do not give consistent answers across contexts are problematic when used for tasks with expectations of consistency–e.g. question-answering, explanations, etc. Our work presents an evaluation benchmark for self-consistency in cases of under-specification where two or more answers can be correct. We conduct a series of behavioral experiments on the OpenAI model suite using an ambiguous integer sequence completion task. We find that average consistency ranges from 67% to 82%, far higher than would be predicted if a model’s consistency was random, and increases as model capability improves. Furthermore, we show that models tend to maintain self-consistency across a series of robustness checks, including prompting speaker changes and sequence length changes. These results suggest that self-consistency arises as an emergent capability without specifically training for it. Despite this, we find that models are uncalibrated when judging their own consistency, with models displaying both over- and under-confidence. We also propose a nonparametric test for determining from token output distribution whether a model assigns non-trivial probability to alternative answers. Using this test, we find that despite increases in self-consistency, models usually place significant weight on alternative, inconsistent answers. This distribution of probability mass provides evidence that even highly self-consistent models internally compute multiple possible responses.

Self-Consistency of Large Language Models under Ambiguity

Empowering the survivor of head injury presentation is a model for collaborative treatment including a resource coordinator, occupational and physical therapy working primarily with adults. Motivational interviewing is program foundation. The treatment is individualized to build on the client’s strengths with all clinicians educating in core skills of recovery.

Empowering the Survivor of Head Injury: Collaborative treatment model

Potential Use of Peripheral Nerve Grafts to Treat CNS Diseases

Understanding Respiratory Muscle Strength Training: Evidence and Clinical Implications for Speech-Language Pathologists

Musculoskeletal disorders effect one in every two Americans costing $213 billion a year as the #1 cause of disability worldwide, with cost and disability rates of musculoskeletal disorders (MSDs) continues to escalate. Speakers will discuss new ways to reduce this workforce epidemic through cost effective physical therapy (PT) and virtual care from employer and employee perspectives.

Efficiency & Outcomes of Direct Access to Physical Therapy for Musculoskeletal Disorders vs Physician-First Access in the US

Teletherapeutic devices enable new possibilities to track the therapy process by gathering and visualizing training statistics. For the therapist this can mean a quicker but also a deeper look at the therapy results. For the patient, the right presentation of data can foster the motivation and adherence to the therapy. In an iterative user centered design process, requirements of data visualization were gathered and the developed solution was validated with health professionals. For the therapists, the most important part is to have a comprehensive overview of conducted exercises and training duration. For the patient, the visualization should be simple and motivative. Further validation, especially with patients is needed.

Underline digital video library

"Honey, Tell Me What's Wrong'', Global Explanation of Textual Discriminative Models through Cooperative Generation

Self-Consistency of Large Language Models under Ambiguity

Empowering the Survivor of Head Injury: Collaborative treatment model

Potential Use of Peripheral Nerve Grafts to Treat CNS Diseases

Understanding Respiratory Muscle Strength Training: Evidence and Clinical Implications for Speech-Language Pathologists

Efficiency & Outcomes of Direct Access to Physical Therapy for Musculoskeletal Disorders vs Physician-First Access in the US

Telerehabilitation Platform REHA2030: Visualization of training statistics for patient and therapist

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES