CONCEPT
Post-Training Specialization
The final stages of model training — supervised fine-tuning, RLHF, DPO, preference optimization — can be tuned to improve a specific benchmark's scores without improving the underlying capability, a structural form of Goodhart's Law built into the optimization pipeline.
Post-training is the sequence of fine-tuning stages applied after pre-training ends: supervised fine-tuning on curated conversations, reinforcement learning from human feedback, direct preference optimization, and related methods. Each stage shapes how the model responds to prompts. When labs face commercial pressure to publish high benchmark scores, post-training becomes the stage at which benchmark-specific optimization occurs most naturally — and most invisibly. Unlike training-data contamination, which is an accident of open-web pre-training, post-training specialization is deliberate, though rarely advertised. It is the primary modern mechanism by which reported scores come to exceed real-world deployment performance.
In The You On AI Field Guide
A benchmark like MMLU consists of multiple-choice questions with four options. A model can be post-trained to format its responses in MMLU's expected style, to select its answer with the confidence patterns MMLU's grading expects, to avoid the hedging and refusal patterns that penalize it on MMLU, and to structure its chain-of-thought