Agentic Discovery with Active Hypothesis Exploration for Visual Recognition

TL;DR

HypoExplore is an agentic framework that formulates neural architecture discovery as hypothesis-driven scientific inquiry, using evolutionary strategies and language models to improve visual recognition models.

Contribution

It introduces a novel hypothesis-driven approach with a Trajectory Tree and Hypothesis Memory Bank for neural architecture search, demonstrating improved performance and transferability.

Findings

Achieved 94.11% accuracy on CIFAR-10, improving from 18.91%.

Discovered architectures that generalize across datasets like CIFAR-100 and Tiny-ImageNet.

Built a framework that enhances understanding of the neural architecture design space.

Abstract

We introduce HypoExplore, an agentic framework that formulates neural architecture discovery for visual recognition as a hypothesis-driven scientific inquiry. Given a human-specified high-level research direction, HypoExplore ideates, implements, evaluates, and improves neural architectures through evolutionary branching. New hypotheses are created using a large language model by selecting a parent hypothesis to build upon, guided by a dual strategy that balances exploiting validated principles with resolving uncertain ones. Our proposed framework maintains a Trajectory Tree that records the lineage of all proposed architectures, and a Hypothesis Memory Bank that actively tracks confidence scores acquired through experimental evidence. After each experiment, multiple feedback agents analyze the results from different perspectives and consolidate their findings into hypothesis confidence updates. Our framework is tested on discovering lightweight vision architectures on CIFAR-10, with the best achieving 94.11% accuracy evolved from a root node baseline that starts at 18.91%, and generalizes to CIFAR-100 and Tiny-ImageNet. We further demonstrate applicability to a specialized domain by conducting independent architecture discovery runs on MedMNIST, which yield a state-of-the-art performance. We show that hypothesis confidence scores grow increasingly predictive as evidence accumulates, and that the learned principles transfer across independent evolutionary lineages, suggesting that HypoExplore not only discovers stronger architectures, but can help build a genuine understanding of the design space.