Together AI· Inference platformResearch paper

Large Reasoning Models Fail to Follow Instructions During Reasoning: A Benchmark Study

This blog post introduces ReasonIF, a benchmark for evaluating instruction-following in reasoning traces of large reasoning models (LRMs). It finds that frontier LRMs such as GPT-OSS-120B, Qwen3-235B, and DeepSeek-R1 fail to follow reasoning instructions more than 75% of the time, and performance degrades with task difficulty.

From the source

We find frontier LRMs, including GPT-OSS-120B, Qwen3-235B, and DeepSeek-R1 fail to follow reasoning instructions more than 75% of time.

together.ai

Announced

22 Oct 2025 00:00 UTC

Detected

03 Aug 2026 21:12 UTC (+285d after)