7 papers
The Reasoning Tax: Token Economics of LLM Reasoning Across Task Types and Deployment Contexts
Sachin Gopal Wani, Ajay Dholakia, David Ellison
Accuracy-only benchmarking of reasoning-capable large language models misses a central deployment question: when do extended thinking tokens earn their cost? We introduce the Token…
KC-Agent: A Dual-Process Cognitive Architecture for Efficient ML Model Improvement
Gusseppe Bravo-Rocca, Jordi Guitart, Ajay Dholakia +2
Data drift poses significant challenges for machine learning systems in production, requiring continuous model updates to maintain performance. We present KC-Agent, a dual-process…
Benchmarking Distilled Language Models: Performance and Efficiency in Resource-Constrained Settings
Sachin Gopal Wani, Eric Page, Ajay Dholakia +1
Knowledge distillation offers a transformative pathway to developing powerful, yet efficient, small language models (SLMs) suitable for resource-constrained environments. In this p…
Feature Engineering for Agents: An Adaptive Cognitive Architecture for Interpretable ML Monitoring
Gusseppe Bravo-Rocca, Peini Liu, Jordi Guitart +3
Monitoring Machine Learning (ML) models in production environments is crucial, yet traditional approaches often yield verbose, low-interpretability outputs that hinder effective de…
TADIL: Task-Agnostic Domain-Incremental Learning through Task-ID Inference using Transformer Nearest-Centroid Embeddings
Gusseppe Bravo-Rocca, Peini Liu, Jordi Guitart +2
Machine Learning (ML) models struggle with data that changes over time or across domains due to factors such as noise, occlusion, illumination, or frequency, unlike humans who can…
Human-in-the-loop online multi-agent approach to increase trustworthiness in ML models through trust scores and data augmentation
Gusseppe Bravo-Rocca, Peini Liu, Jordi Guitart +3
Increasing a ML model accuracy is not enough, we must also increase its trustworthiness. This is an important step for building resilient AI systems for safety-critical application…
Scanflow: A multi-graph framework for Machine Learning workflow management, supervision, and debugging
Gusseppe Bravo-Rocca, Peini Liu, Jordi Guitart +4
Machine Learning (ML) is more than just training models, the whole workflow must be considered. Once deployed, a ML model needs to be watched and constantly supervised and debugged…