Unverified50% confidenceOpinionExact time
The impact of network scale and training strategy on performance is not monotonic; on the 118-bus system, training with simulated states instead drags down accuracy, indicating that transferability and scale extension still require further study.
1
Sources
50%
Confidence
Long-term
Relevance
11/10/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOn the 118-bus system, a network trained with simulated states is instead inferior to directly using this operator (the Physics-Informed Graph Neural Operator), which is faster and more accurate than hard-constrained moving horizon estimation.77% similarUnverifiedA core long-standing challenge for non-backpropagation methods is that their final performance on large-scale tasks is often difficult to match that of mature backpropagation training74% similarUnverifiedMaRN's mapping model training is often much slower than direct training, because the mapping network itself introduces additional forward/backward propagation overhead.74% similarUnverifiedThe premise for small-scale scaling experiments to hold is keeping the training recipe (optimizer, learning rate schedule, batch size, sequence length, data distribution, mixed-precision strategy, etc.) unchanged across different scale points, changing only the parameter count and token count.74% similarUnverifiedCompared to MoE architectures, dense models have more uniform gradient flow during training, are less prone to expert collapse, and exhibit more deterministic inference behavior.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1153659API
curl https://kongchang.com/api/v1/knowledge/claims/1153659MCP
get_claim(id=1153659)