We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Learning to Solve Complex Tasks by Talking to Agents

Abstract: Humans often solve complex problems by interacting (in natural language) with existing agents, such as AI assistants, that can solve simpler sub-tasks. These agents themselves can be powerful systems built using extensive resources and privately held data. In contrast, common NLP benchmarks aim for the development of self-sufficient models for every task. To address this gap and facilitate research towards ``green'' AI systems that build upon existing agents, we propose a new benchmark called CommaQA that contains three kinds of complex reasoning tasks that are designed to be solved by ``talking'' to four agents with different capabilities. We demonstrate that state-of-the-art black-box models, which are unable to leverage existing agents, struggle on CommaQA (exact match score only reaches 40pts) even when given access to the agents' internal knowledge and gold fact supervision. On the other hand, models using gold question decomposition supervision can indeed solve CommaQA to a high accuracy (over 96\% exact match) by learning to utilize the agents. Even these additional supervision models, however, do not solve our compositional generalization test set. Finally the end-goal of learning to solve complex tasks by communicating with existing agents \emph{without relying on any additional supervision} remains unsolved and we hope CommaQA serves as a novel benchmark to enable the development of such systems.
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2110.08542 [cs.CL]
  (or arXiv:2110.08542v1 [cs.CL] for this version)

Submission history

From: Tushar Khot [view email]
[v1] Sat, 16 Oct 2021 10:37:34 GMT (392kb,D)

Link back to: arXiv, form interface, contact.