Setting αk to the maximum gradient norm performs best among tested strategies on NYUv2 (Figure 6).

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• DB-MTL training losses decrease smoothly and gradient norms are lower than EW on NYUv2, indicating training stability.finding0.790
  Training stability analysis.
• Setting aggregated gradient scaling factor to maximum gradient norm performs best for task balancingclaim0.770
  Empirical finding on choice of αk in gradient normalization strategy
• The gradient-magnitude balancing method outperforms GradNorm on NYUv2, Cityscapes, Office-31, Office-Home.finding0.759
  Comparison of gradient-magnitude balancing with GradNorm.
• The proposed gradient-magnitude balancing method consistently outperforms GradNorm, as it guarantees equal gradient magnitudes and considers update magnitude.claim0.755
  Advantage over GradNorm.
• Combining loss-scale and gradient-magnitude balancing achieves Δp = +1.15±0.16 on NYUv2.finding0.752
  Full DB-MTL ablation result.
• If EI maximization is used as a regularization in representation learning, then OOD generalization will improve beyond current invariant risk minimization methods.hypothesis0.751
  Proposed conjecture in §4.3.1.
• Software implementations for all of the models/behaviours presented are common for n = 2, and can be made very efficient for α_i that map many objects onto a much smaller set of object families.claim0.749
  Claim about current practical feasibility and efficiency of 2-way associative implementations.
• DB-MTL has similar per-epoch running time to gradient balancing methods on NYUv2, slower than loss balancing methods.finding0.745
  Computational efficiency comparison.