A model tops a published general benchmark. A team adopts it for extracting fields from their own scanned forms and finds it performs worse than the smaller model they were using. What does this illustrate?

• AThe benchmark result must have been fabricated.
• BBenchmark scores measure general capability and correlate weakly with narrow applied tasks.
• CThe larger model needs fine tuning before it can be compared fairly.
• DThe team should raise temperature to unlock the larger model's capability.