SciIF: Benchmarking Scientific Instruction Following Towards Rigorous Scientific Intelligence

TL;DR

SciIF introduces a comprehensive benchmark to evaluate large language models on their ability to follow scientific instructions rigorously, emphasizing constraint adherence, evidence provision, and reasoning transparency across multiple scientific disciplines.

Contribution

The paper presents SciIF, a novel benchmark that assesses scientific instruction following with emphasis on constraint compliance and evidence, addressing limitations of existing evaluation standards.

Findings

Models can be evaluated on both correctness and constraint adherence.

SciIF enables diagnosis of reasoning failures in scientific problem-solving.

Benchmark promotes development of more reliable scientific AI agents.

Abstract

As large language models (LLMs) transition from general knowledge retrieval to complex scientific discovery, their evaluation standards must also incorporate the rigorous norms of scientific inquiry. Existing benchmarks exhibit a critical blind spot: general instruction-following metrics focus on superficial formatting, while domain-specific scientific benchmarks assess only final-answer correctness, often rewarding models that arrive at the right result with the wrong reasons. To address this gap, we introduce scientific instruction following: the capability to solve problems while strictly adhering to the constraints that establish scientific validity. Specifically, we introduce SciIF, a multi-discipline benchmark that evaluates this capability by pairing university-level problems with a fixed catalog of constraints across three pillars: scientific conditions (e.g., boundary checks and assumptions), semantic stability (e.g., unit and symbol conventions), and specific processes(e.g., required numerical methods). Uniquely, SciIF emphasizes auditability, requiring models to provide explicit evidence of constraint satisfaction rather than implicit compliance. By measuring both solution correctness and multi-constraint adherence, SciIF enables finegrained diagnosis of compositional reasoning failures, ensuring that LLMs can function as reliable agents within the strict logical frameworks of science.