CD-RFT defines a Post-training Control Coefficient separating activation from control, and shows that regularizing a new Shared Control Bottleneck improves multi-task pass@1 and pass@k over matched GRPO.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Control-Diverse Reinforcement Fine-Tuning: Decoupling the Shared Control Bottleneck of RL Post-Training
CD-RFT defines a Post-training Control Coefficient separating activation from control, and shows that regularizing a new Shared Control Bottleneck improves multi-task pass@1 and pass@k over matched GRPO.