Beyond Binary Success: A Diagnostic Meta-Evaluation Framework for Fine-Grained Manipulation

TL;DR

MetaFine is a diagnostic framework that dissects fine-grained manipulation skills into understanding, perception, and controlled behavior, revealing hidden model weaknesses and guiding targeted improvements.

Contribution

The paper introduces MetaFine, a novel diagnostic evaluation framework that reconstructs heterogeneous benchmarks into diagnostic scenarios, exposing model failures and guiding targeted enhancements.

Findings

Visual encoder quality is a key bottleneck for fine-grained manipulation.

MetaFine reveals severe dimension-specific failures in state-of-the-art models.

Hybrid real-sim validation improves physical benchmarking stability.

Abstract

Fine-grained manipulation marks a regime where global scene context no longer suffices, and success hinges on the tight coupling of local attribute grounding, high-fidelity spatial perception, and constraint-respecting motor execution. However, current embodied AI benchmarks collapse these capacities into binary success rates, systematically inflating reported capabilities by up to 70% and masking the architectural bottlenecks that impede real-world deployment. We introduce MetaFine, a diagnostic meta-evaluation framework that disentangles manipulation competency along three axes: understanding, perception, and controlled behavior. Built on a compositional task graph, MetaFine absorbs heterogeneous external benchmarks and reconstructs them into diagnostic scenarios of varying complexity under a unified protocol. Evaluating state-of-the-art vision-language-action (VLA) models through this lens exposes severe dimension-specific failures invisible to conventional metrics. Through targeted causal intervention, we identify the visual encoder's ability to preserve local spatial structure as a key bottleneck for fine-grained precision: improving it directly unlocks previously inaccessible manipulation capabilities without modifying downstream policies. MetaFine further supports hybrid real-sim validation, using limited paired real-world rollouts to calibrate scalable simulation-based estimates for more stable physical benchmarking. By shifting evaluation from ranking to diagnosis, MetaFine turns benchmarking into an actionable compass for repairing the layered capacities underlying genuine physical dexterity. The MetaFine framework, benchmarks, and supporting resources will be publicly released at our project page: https://metafine.github.io/.