Unverified50% confidenceTradeoffExact time
For classification tasks with discrete and closed label spaces, the training signal from SFT is already clear enough, and introducing reinforcement learning would instead increase the difficulty of hyperparameter tuning and training instability
1
Sources
50%
Confidence
Long-term
Relevance
10/10/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn the post-training phase, Motif adopts a "failure-driven SFT" strategy: identifying model weaknesses through benchmark testing and internal scenarios, then generating targeted datasets for training.76% similarVerifiedDuring SFT training, the model only computes loss on the assistant portion (loss masking technique), meaning it only learns how to answer, not how to ask questions.75% similarVerifiedLarge model training goes through three core stages: pre-training, supervised fine-tuning (SFT), and preference alignment.75% similarUnverifiedPre-training and SFT provide only positive supervision, whereas DPO and adjacent RL techniques provide the leverage of negative supervision, which can reduce the probability of erroneous sequences.74% similarVerifiedThe train-serve consistency problem refers to logical discrepancies between offline feature computation during training and online real-time feature computation during serving, which can cause the model's online performance to be severely worse than its offline evaluation.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/994898API
curl https://kongchang.com/api/v1/knowledge/claims/994898MCP
get_claim(id=994898)