Search papers, labs, and topics across Lattice.
This paper introduces PredicateLongBench, a benchmark designed to systematically evaluate long-context reasoning in large language models (LLMs) by measuring performance along various axes of difficulty. The authors highlight that existing benchmarks often focus on average-case performance and lack robustness, which limits their effectiveness in understanding model limitations. Through two generation pipelines鈥攐ne synthetic and one real-world鈥攖he study reveals that state-of-the-art models struggle with increasing task difficulty, emphasizing the need for better evaluation methods in long-context tasks.
Frontier models falter as task difficulty scales, revealing critical gaps in long-context reasoning capabilities.
Large language models (LLMs) have demonstrated rapidly improving long-context capabilities, prompting a wave of benchmarks designed to evaluate them. However, existing long-context evaluations - from Needle-in-a-Haystack (NIAH) tests to more recent multi-hop reasoning and summarization tasks - predominantly measure average-case performance, and many are either saturated or lack robustness. Notably absent is a systematic way to probe how models perform as we scale up the difficulty of tasks along various axes. We address this gap by proposing PredicateLongBench, a benchmark that stress-tests long-context reasoning by asking models to identify the longest contiguous subsequence of words in a long input that satisfies given predicates/constraints (e.g., lexicographic ordering), drawn from a broader predicate class. The central innovation of our benchmark is the identification and systematic exploration of multiple different axes of difficulty which test multiple aspects of long context understanding. We provide two complementary generation pipelines - a fully synthetic setup using random word-like strings, and a real-world setup that samples words from natural documents while preserving their distributional properties. We find that frontier models struggle to perform well as we scale up the difficulty of tasks along our axes, demonstrating the utility of our benchmark in understanding the limitations of current long-context capabilities. Furthermore, the tasks in PredicateLongBench, though challenging, are conceptually simple and do not require LLM-based generations or judges.
