Unverified50% confidenceBenchmarkExact time
On the HybridCodeAuthorship external benchmark (no training set), the method achieved a zero-shot line-level F1 of 62.5, exceeding the best line-level result of 56 reported in the original paper.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/21/2026
First Seen
Valid until: 1/19/2027
Sources
Related Entities
Related Claims
UnverifiedThe Label-Only method achieves an accuracy of 0.896 and a span F1 of 0.564, with verbatim citation improving from 0.597 before training to 0.729.75% similarUnverifiedThe Macro F1 of the factor identification module is only 0.5562, significantly lower than the risk assessment and evidence tracing tasks.68% similarUnverifiedThe milestones publicly demonstrated by VKORE include: training a 46-million-parameter model from scratch, running on 32 inference-only NPU chips, with a loss correlation of 0.99 with a GPU training baseline.67% similarUnverifiedThe best version of the activation-based representation method achieves a correlation coefficient of 0.45 with the generalization matrix, while the description-based baseline method achieves only 0.05.66% similarUnverifiedIn the multi-channel forging process case study, ES-CP achieved a macro F1 score of 0.8266% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1079469API
curl https://kongchang.com/api/v1/knowledge/claims/1079469MCP
get_claim(id=1079469)