RPSFT improves the in-domain versus out-of-domain performance trade-off during LLM supervised fine-tuning by penalizing rotations in pretrained singular subspaces as a proxy for loss-sensitive directions.
Rupam Mahmood
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
verdicts
UNVERDICTED 4roles
method 1polarities
use method 1representative citing papers
Clipped least-squares importance fitting enables weighted conformal prediction to achieve dataset-conditional coverage guarantees under unbounded covariate shifts by bounding undercoverage and estimating a corrective inflation factor from data.
FADE adapts per-parameter weight decay rates online via approximate meta-gradient descent to improve controlled forgetting over fixed decay in online tracking and streaming classification.
Survey unifies the definition of plasticity loss in DRL, taxonomizes over 50 mitigations, identifies evaluation gaps, and finds general regularization often outperforms domain-specific methods.
citing papers explorer
-
Rotation-Preserving Supervised Fine-Tuning
RPSFT improves the in-domain versus out-of-domain performance trade-off during LLM supervised fine-tuning by penalizing rotations in pretrained singular subspaces as a proxy for loss-sensitive directions.
-
Weight Clipping for Robust Conformal Inference under Unbounded Covariate Shifts
Clipped least-squares importance fitting enables weighted conformal prediction to achieve dataset-conditional coverage guarantees under unbounded covariate shifts by bounding undercoverage and estimating a corrective inflation factor from data.
-
Learning to Forget: Continual Learning with Adaptive Weight Decay
FADE adapts per-parameter weight decay rates online via approximate meta-gradient descent to improve controlled forgetting over fixed decay in online tracking and streaming classification.
-
Plasticity Loss in Deep Reinforcement Learning: A Survey
Survey unifies the definition of plasticity loss in DRL, taxonomizes over 50 mitigations, identifies evaluation gaps, and finds general regularization often outperforms domain-specific methods.