Results

Results

Rank	Team	Final Score	Phase 1 -	Phase 1 -	Phase 2 -	Dataset Size*	Duplication Ratio**
1	VeriBest	0.769	49.97	46.86	32.30	529,756	0.23
2	SDUAES	0.657	44.51	54.18	41.75	10,042	0.45
3	UMD	0.637	44.36	52.74	39.06	24,886	0.63
4	lcm-verilog	0.544	41.13		39.51	10,318	0.80
5	DRIVE	0.533	42.44	53.52	37.21	33,160	0.40
6	Origen	0.523	49.75	48.54	24.77	168,554	0.24
7	CSDL	0.475	39.71		37.68	4,003	0.64
8	nathan	0.403	39.38		34.99	82	0.20
9	CircuitUp	0.359	39.75		32.90	151	0.95
10	UCLA VAST lab	0.356	42.01	51.95	28.11	2,469	0.81
11	BisonCAD	0.344	35.36		35.36	15,009	0.36
12	Chip Synapses	0.332	39.76		31.80	26,628	0.30
13	LLM4HW	0.197	41.55	51.51		1,876	0.01
14	Aurangabad	0.145	41.83	53.18		238	0.14
15	HDLOJ	0.099	41.17		21.32	4,719	0.74
16	UFO	0.088	40.49			2	0.99
17	Team Flames	0.085	40.35		0.53	1,560	0.26

* Dataset size refers to the number of unique code samples in the submitted dataset after deduplication within the submitted dataset itself as well as against VerilogEval, RTLLM, and MG-Verilog.
** Duplication ratio is defined as the ratio of duplicated code samples to unique code samples in the dataset, calculated by comparing against VerilogEval, RTLLM, MG-Verilog, and the corresponding participant's submitted dataset.