A Benchmark for Deep Information Synthesis

TL;DR

DEEPSYNTH is a new benchmark designed to evaluate large language models on complex, real-world information synthesis tasks across multiple domains, revealing current limitations in reasoning and hallucination management.

Contribution

The paper introduces DEEPSYNTH, a comprehensive benchmark with 120 tasks across diverse domains, specifically crafted to assess LLMs' abilities in information synthesis and reasoning.

Findings

State-of-the-art LLMs achieve low scores, indicating high difficulty.

Current agents struggle with hallucinations and reasoning over large data.

DEEPSYNTH highlights the need for improved reasoning capabilities in LLMs.

Abstract

Large language model (LLM)-based agents are increasingly used to solve complex tasks involving tool use, such as web browsing, code execution, and data analysis. However, current evaluation benchmarks do not adequately assess their ability to solve real-world tasks that require synthesizing information from multiple sources and inferring insights beyond simple fact retrieval. To address this, we introduce DEEPSYNTH, a novel benchmark designed to evaluate agents on realistic, time-consuming problems that combine information gathering, synthesis, and structured reasoning to produce insights. DEEPSYNTH contains 120 tasks collected across 7 domains and data sources covering 67 countries. DEEPSYNTH is constructed using a multi-stage data collection pipeline that requires annotators to collect official data sources, create hypotheses, perform manual analysis, and design tasks with verifiable answers. When evaluated on DEEPSYNTH, 11 state-of-the-art LLMs and deep research agents achieve a maximum F1 score of 8.97 and 17.5 on the LLM-judge metric, underscoring the difficulty of the benchmark. Our analysis reveals that current agents struggle with hallucinations and reasoning over large information spaces, highlighting DEEPSYNTH as a crucial benchmark for guiding future research.