Unverified50% confidenceFactExact time
Supervised fine-tuning (SFT) optimizes imitation loss rather than a behavioral safety objective, so it readily exhibits the phenomenon of "working within the training distribution but failing out of distribution."
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedDuring 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.80% similarUnverifiedFor 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 instability80% 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.78% similarUnverifiedSFT is usually much cheaper than pretraining but is extremely sensitive to data quality; data ratios, the number of epochs, and the learning rate all significantly affect results.77% similarUnverifiedIn 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% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1102333API
curl https://kongchang.com/api/v1/knowledge/claims/1102333MCP
get_claim(id=1102333)