Skip to main content
Search

# Gradient Alignment

2 related articles

Subscribe to updates
  1. arXiv

    Self-distillation's reasoning gains don't survive new problems: coding help stays inside the training set, and gradients often point away from policy gradients

    Two self-distillation methods lose or reverse their reported gains on unseen problems across code, science, tool use and math, and the math improvement comes mainly from transferring thinking-mode behavior rather than from privileged information such as reference solutions.

    Massachusetts Institute of Technology
    Jason Ken Adhinarta（First author）
    Self-DistillationSelf-Distillation Policy OptimizationOn-Policy Self-DistillationGRPOGradient Alignment
  2. arXiv

    OmniTaskonomy maps 19 generation tasks against 25 understanding capabilities: I2I-then-I2T training improves understanding, and gradient alignment tracks the gains

    Using controlled paired image-to-image (I2I) generation and image-to-text (I2T) understanding tasks on the unified multimodal model BAGEL, the authors find that a recipe of I2I training that updates shared parameters followed by I2T finetuning improves understanding with gains that grow as I2I data increases, build OmniTaskonomy, a taxonomy spanning 19 I2I tasks and 25 understanding capabilities, produce a selective task-dependent generation-to-understanding transfer map, and find that stronger gradient alignment between generation and understanding is associated with larger downstream transfer gains.

    Jiaxin Ge（First author）
    OmniTaskonomyCross-Modal TransferGradient AlignmentBAGELUnified Multimodal Models

You’re all caught up