Unverified50% confidenceBenchmarkExact time
The paper evaluates on two benchmarks, LoCoMo and MemGallery, using four open-source and closed-source backbone models, improving macro-F1 by 2.23 to 3.14 points.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/28/2026
First Seen
Valid until: 1/26/2027
Sources
Related Entities
Related Claims
UnverifiedUnder four-view test-time augmentation, ensembling the flat model with the label-correlation model achieved the best configuration on six of eight metrics: exact match 0.694, micro-F1 0.678, macro-F1 0.456, and macro-AUC 0.972.67% similarUnverifiedThe state-space-model-based multi-scale classifier proposed in the paper achieved a macro F1 of 0.894 and a Gleason 5 F1 of 0.802 on the PANDA dataset using 5-fold cross-validation.66% similarUnverifiedIn the BioPhys-Bridge evaluation, DeepSeek-V4-Flash achieved the highest score with an evidence ID F1 score of 0.360.65% similarUnverifiedCompared with the best single-scale model, this method achieves a 12% relative improvement in macro F1 and a 21% relative improvement in Gleason 5 F1 on the PANDA dataset.64% similarUnverifiedGLM-5.3 achieved a state-of-the-art score among open-source models on the Agents' Last Exam benchmark.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1106327API
curl https://kongchang.com/api/v1/knowledge/claims/1106327MCP
get_claim(id=1106327)