Unverified50% confidenceBenchmarkExact time
Fiorillo v0.5 achieves a macro-F1 of 0.9248, compared to the baseline's 0.8668.
1
Sources
50%
Confidence
Long-term
Relevance
10/7/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe new system's F1 score improved from 0.606 to 0.702.70% similarVerifiedMacro-F1 calculates F1 for each class separately and then takes the arithmetic mean, while Micro-F1 aggregates the TP/FP/FN across all classes first and then computes F1 uniformly.63% similarUnverifiedUsing a quantization combination of INT8 weights with INT4 embedding layers in ONNX, the F1 score difference is only approximately 0.005 compared to more conservatively quantized versions.63% similarUnverifiedThe F100 inherited the F5's Multi-CAM1300 autofocus system and 3D matrix metering, with a body weight of approximately 785 grams and a price roughly half that of the F5.60% similarUnverifiedFiorillo v0.5 uses Qwen3-4B-Base as its backbone, with a low-rank adapter (LoRA) and a dedicated decision head stacked on top, fine-tuned exclusively for the EI task.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/983040API
curl https://kongchang.com/api/v1/knowledge/claims/983040MCP
get_claim(id=983040)