Unverified50% confidenceFactExact time
The experiments focus on discrete groups such as S4 and sparse graphs; extension to larger scales, continuous groups, and dense graphs has not yet been carried out.
1
Sources
50%
Confidence
Long-term
Relevance
10/28/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedResearch shows that large-scale multimodal models trained on massive data have to some extent developed emergent compositional generalization capabilities, but still fail on certain extreme combinations.62% similarUnverifiedWhen diffusion models generate large-scale crowds, individual faces occupy extremely low resolution, making it difficult for the model to maintain clear and natural facial features within small regions.61% similarUnverifiedLeslie矩阵的主特征值（谱半径）决定种群长期增长率：大于1则增长，小于1则趋于灭绝59% similarUnverifiedThe research shows that capability improvements may lower the critical threshold for population takeoff, so the critical population size needs to be re-estimated for each new generation of models.59% similarUnverifiedThe study compared three adaptation strategies: full fine-tuning, LoRA low-rank adaptation, and post-hoc weight-space merging.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1106044API
curl https://kongchang.com/api/v1/knowledge/claims/1106044MCP
get_claim(id=1106044)