A Subgoal-driven Framework for Improving Long-Horizon LLM Agents

Taiyi Wang, Sian Gooding, Florian Hartmann, Oriana Riva, Edward Grefenstette · 2026

Read A Subgoal-driven Framework for Improving Long-Horizon LLM Agents by Taiyi Wang, Sian Gooding, Florian Hartmann, Oriana Riva, Edward Grefenstette (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-19685.txt.

Description

Wang and co-authors present a subgoal-driven framework that improves long-horizon language model agents. Agents on web and app tasks often lose track as new information arrives. Their online planner uses strong proprietary models to split goals into subgoals. A second piece, MiRA, trains agents with dense milestone-based rewards. On WebArena-Lite, the planner lifts Gemini success by about ten points. MiRA raises an open Gemma model from about 6 percent to 43 percent success. That beats several proprietary systems and the prior open-model best.

Catalog record

CREATORTaiyi Wang, Sian Gooding, Florian Hartmann, Oriana Riva, Edward Grefenstette
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE136559
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.19685

Browse the full library · About the Tower of Babel