aiqbench analysis
ATHENA introduces an efficient influence function framework specifically designed for scaling data curation in billion-parameter Vision-Language-Action models across multiple tasks. By leveraging Kronecker structures and random truncated approximations, the method achieves a 313x speedup while maintaining performance comparable to full-data fine-tuning on both simulation and real robots. This advancement addresses critical computational bottlenecks in training large-scale robotic agents, making high-quality data selection feasible for complex multitask environments.
ATHENA: Accelerated Multi-Task Heterogeneous Influence Functions for Robot Data Curation
arXiv:2606.16208v1 Announce Type: new Abstract: In robot imitation learning, influence functions provide a principled approach to quantify each demonstration's effect on robot task outcomes, yet scaling them to billion-parameter Vision-Language-Action (VLA) models is limited by computational and multitask bottlenecks. To this end, we propose ATHENA, an influence function framework tailored for multitask VLA data curation at a billion-parameter scale. Concretely, it leverages the Kronecker structure of linear-layer gradients to reduce projection cost, and approximates dense Hessian inversion with a rank-r Random Truncated Approximation, achieving about a 313.4x speedup in influence computation. Furthermore, ATHENA formulates global and local interactive influence to balance data curation across 50 jointly trained tasks. Extensive evaluations on RoboTwin 2.0 and real-robot deployment, covering 9.34 and 6.90 hours of demonstrations, respectively, show that ATHENA matches or exceeds full-data joint fine-tuning using only 50% of demonstrations in simulation and 66.7% of data across six real-robot tasks. Overall, ATHENA demonstrates its effectiveness for data curation in billion-parameter multitask VLA fine-tuning.
Abstract:In robot imitation learning, influence functions provide a principled approach to quantify each demonstration's effect on robot task outcomes, yet scaling them to billion-parameter Vision-Language-Action (VLA) models is limited by computational and multitask bottlenecks. To this end, we propose ATHENA, an influence function framework tailored for multitask VLA data curation at a billion-parameter scale. Concretely, it leverages the Kronecker structure of linear-layer gradients to reduce projection cost, and approximates dense Hessian inversion with a rank-r Random Truncated Approximation, achieving about a 313.4x speedup in influence computation. Furthermore, ATHENA formulates global and local interactive influence to balance data curation across 50 jointly trained tasks. Extensive evaluations on RoboTwin 2.0 and real-robot deployment, covering 9.34 and 6.90 hours of demonstrations, respectively, show that ATHENA matches or exceeds full-data joint fine-tuning using only 50% of demonstrations in simulation and 66.7% of data across six real-robot tasks. Overall, ATHENA demonstrates its effectiveness for data curation in billion-parameter multitask VLA fine-tuning.
Robotics (cs.RO)
arXiv:2606.16208 [cs.RO]
(or arXiv:2606.16208v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2606.16208 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 249 words
Related coverage
Terastal: Layer-Variant-based Scheduling for Real-Time Multi-DNN Workloads on Heterogeneous Accelerators
arXiv · Jun 8, 2026
From Legacy Fortran to Portable Kokkos: An Autonomous Agentic AI Workflow
arXiv · Oct 29, 2026
PrismServe: Efficient LLM Serving on Heterogeneous Accelerator Clusters
arXiv · Nov 4, 2026
From Inference Engine to Inference Control Plane: Connecting vLLM, llm-d, and the Evolution of Efficient Distributed LLM Serving
arXiv · Sep 22, 2026