Dataset Viewer
Auto-converted to Parquet Duplicate
Unnamed: 0
stringlengths
11
22
R-Syn-1
stringlengths
14
14
R-Syn-Max
stringlengths
14
14
R-Sem
stringlengths
14
14
S-Syn-1
stringlengths
14
14
S-Syn-Max
stringlengths
14
14
S-Sem-R
stringlengths
14
14
S-Sem-W-1
stringlengths
14
14
S-Sem-W-max
stringlengths
14
14
total
stringlengths
14
14
Qwen-3.5-397B
0.941 (±0.215)
0.994 (±0.015)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.940 (±0.211)
0.940 (±0.211)
0.977 (±0.133)
Claude Opus 4.6
0.934 (±0.234)
0.996 (±0.013)
0.657 (±0.415)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.747 (±0.362)
0.837 (±0.311)
0.896 (±0.270)
Claude Sonnet 4.6
0.934 (±0.234)
0.996 (±0.013)
0.282 (±0.300)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.747 (±0.362)
0.797 (±0.338)
0.844 (±0.321)
Gemini 3 Flash Preview
0.961 (±0.168)
0.995 (±0.014)
0.973 (±0.061)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.760 (±0.367)
0.850 (±0.312)
0.942 (±0.200)
GPT5.2-chat
0.962 (±0.168)
0.995 (±0.014)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.807 (±0.332)
0.877 (±0.276)
0.955 (±0.178)
GPT5.4 2026/03
0.934 (±0.234)
0.996 (±0.013)
0.998 (±0.020)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.647 (±0.388)
0.687 (±0.381)
0.908 (±0.253)
Claude 3.5 Haiku
0.937 (±0.203)
0.984 (±0.029)
0.779 (±0.395)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.817 (±0.325)
0.887 (±0.266)
0.925 (±0.232)
Claude 3.5 Sonnet
0.950 (±0.175)
0.990 (±0.022)
0.832 (±0.370)
0.980 (±0.125)
1.000 (±0.000)
1.000 (±0.000)
0.857 (±0.295)
0.857 (±0.295)
0.933 (±0.222)
Deepseek-Coder-33B
0.773 (±0.366)
0.882 (±0.269)
0.263 (±0.297)
0.943 (±0.221)
0.984 (±0.112)
0.313 (±0.309)
0.689 (±0.379)
0.703 (±0.374)
0.694 (±0.396)
Deepseek-R1
0.955 (±0.174)
0.991 (±0.020)
0.992 (±0.089)
0.935 (±0.247)
1.000 (±0.000)
1.000 (±0.000)
0.746 (±0.382)
0.832 (±0.316)
0.931 (±0.226)
Deepseek-Chat-v3
0.843 (±0.347)
0.991 (±0.020)
0.591 (±0.466)
0.957 (±0.202)
0.997 (±0.050)
0.923 (±0.214)
0.702 (±0.384)
0.782 (±0.348)
0.848 (±0.326)
Gemini 1.5 Flash
0.920 (±0.242)
0.983 (±0.028)
0.878 (±0.325)
0.865 (±0.324)
0.910 (±0.272)
1.000 (±0.000)
0.850 (±0.304)
0.850 (±0.304)
0.907 (±0.263)
Gemini 1.5 Pro
0.887 (±0.291)
0.966 (±0.127)
0.796 (±0.399)
0.845 (±0.339)
0.905 (±0.286)
1.000 (±0.000)
0.883 (±0.276)
0.883 (±0.276)
0.896 (±0.282)
Gemini 2.0 Flash Exp
0.986 (±0.025)
0.988 (±0.024)
0.931 (±0.197)
0.994 (±0.079)
1.000 (±0.000)
1.000 (±0.000)
0.604 (±0.394)
0.657 (±0.387)
0.895 (±0.260)
Llama-3.1-70B
0.908 (±0.234)
0.973 (±0.034)
0.559 (±0.484)
0.997 (±0.050)
0.997 (±0.050)
1.000 (±0.000)
0.694 (±0.381)
0.754 (±0.361)
0.860 (±0.311)
Llama-3.1-8B
0.779 (±0.375)
0.915 (±0.228)
0.462 (±0.421)
0.401 (±0.475)
0.521 (±0.477)
0.535 (±0.377)
0.273 (±0.401)
0.355 (±0.425)
0.530 (±0.452)
Llama-3.2-1B
0.250 (±0.366)
0.411 (±0.409)
0.159 (±0.254)
0.026 (±0.143)
0.079 (±0.260)
0.021 (±0.070)
0.010 (±0.050)
0.027 (±0.073)
0.123 (±0.276)
Llama-3.2-3B
0.402 (±0.452)
0.773 (±0.332)
0.344 (±0.397)
0.196 (±0.374)
0.322 (±0.444)
0.308 (±0.373)
0.120 (±0.256)
0.212 (±0.308)
0.335 (±0.416)
Llama-3.3-70B
0.975 (±0.032)
0.978 (±0.029)
0.595 (±0.487)
0.985 (±0.122)
1.000 (±0.000)
1.000 (±0.000)
0.617 (±0.398)
0.671 (±0.385)
0.853 (±0.318)
Llama-3.0-70B
0.961 (±0.114)
0.974 (±0.033)
0.523 (±0.480)
0.955 (±0.208)
0.990 (±0.099)
1.000 (±0.000)
0.645 (±0.405)
0.731 (±0.367)
0.847 (±0.324)
Llama-3.0-8B
0.586 (±0.426)
0.632 (±0.416)
0.219 (±0.290)
0.271 (±0.445)
0.425 (±0.488)
0.615 (±0.337)
0.281 (±0.397)
0.445 (±0.417)
0.434 (±0.435)
Llama-4-Maverick
0.870 (±0.241)
0.974 (±0.033)
0.655 (±0.465)
0.960 (±0.196)
1.000 (±0.000)
0.910 (±0.244)
0.687 (±0.381)
0.815 (±0.327)
0.859 (±0.305)
GPT3.5 2024/01
0.975 (±0.126)
0.995 (±0.014)
0.411 (±0.442)
0.944 (±0.230)
1.000 (±0.000)
0.696 (±0.374)
0.674 (±0.387)
0.707 (±0.376)
0.800 (±0.356)
GPT4o 2024/11
0.937 (±0.212)
0.986 (±0.024)
0.726 (±0.377)
1.000 (±0.000)
1.000 (±0.000)
0.881 (±0.183)
0.817 (±0.325)
0.867 (±0.286)
0.902 (±0.244)
GPT4o-mini 2024/07
0.919 (±0.232)
0.983 (±0.030)
0.384 (±0.415)
0.921 (±0.246)
0.960 (±0.174)
0.962 (±0.089)
0.709 (±0.385)
0.777 (±0.349)
0.827 (±0.333)
GPTo1-mini 2024/09
0.835 (±0.351)
0.992 (±0.018)
0.994 (±0.031)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.697 (±0.379)
0.767 (±0.354)
0.911 (±0.251)
GPTo1-pre 2024/09
0.911 (±0.256)
0.992 (±0.020)
0.658 (±0.373)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.742 (±0.365)
0.812 (±0.329)
0.889 (±0.268)
OpenCoder-8B
0.746 (±0.405)
0.817 (±0.354)
0.167 (±0.285)
0.622 (±0.482)
0.737 (±0.437)
0.400 (±0.422)
0.459 (±0.422)
0.509 (±0.417)
0.557 (±0.454)
Phi-3.5-mini
0.608 (±0.412)
0.639 (±0.390)
0.176 (±0.297)
0.637 (±0.466)
0.683 (±0.450)
0.450 (±0.381)
0.309 (±0.367)
0.350 (±0.371)
0.481 (±0.432)
Phi-3.5-MoE
0.831 (±0.296)
0.841 (±0.287)
0.517 (±0.421)
0.808 (±0.391)
0.932 (±0.238)
0.688 (±0.185)
0.637 (±0.394)
0.648 (±0.389)
0.738 (±0.359)
Phi-3.0-medium-128k
0.838 (±0.318)
0.886 (±0.257)
0.248 (±0.364)
0.547 (±0.475)
0.603 (±0.466)
0.625 (±0.316)
0.360 (±0.412)
0.385 (±0.417)
0.561 (±0.439)
Phi-3.0-mini-128k
0.582 (±0.424)
0.660 (±0.388)
0.263 (±0.333)
0.486 (±0.479)
0.549 (±0.480)
0.428 (±0.336)
0.231 (±0.291)
0.245 (±0.292)
0.431 (±0.415)
Phi-3.0-small-128k
0.346 (±0.394)
0.432 (±0.385)
0.284 (±0.364)
0.366 (±0.427)
0.394 (±0.429)
0.593 (±0.487)
0.278 (±0.352)
0.300 (±0.356)
0.374 (±0.413)
Qwen-2.0-0.5B
0.068 (±0.159)
0.076 (±0.171)
0.085 (±0.205)
0.005 (±0.071)
0.012 (±0.111)
0.040 (±0.136)
0.006 (±0.072)
0.010 (±0.080)
0.038 (±0.138)
Qwen-2.0-1.5B
0.126 (±0.294)
0.145 (±0.314)
0.222 (±0.349)
0.293 (±0.448)
0.351 (±0.465)
0.154 (±0.196)
0.105 (±0.177)
0.115 (±0.182)
0.189 (±0.332)
Qwen-2.5-0.5B
0.053 (±0.153)
0.101 (±0.235)
0.083 (±0.199)
0.157 (±0.360)
0.185 (±0.384)
0.071 (±0.179)
0.061 (±0.128)
0.064 (±0.129)
0.097 (±0.244)
Qwen-2.5-14B
0.781 (±0.393)
0.922 (±0.245)
0.331 (±0.432)
0.897 (±0.303)
0.910 (±0.286)
0.933 (±0.240)
0.658 (±0.378)
0.671 (±0.374)
0.763 (±0.390)
Qwen-2.5-1.5B
0.470 (±0.459)
0.584 (±0.452)
0.266 (±0.339)
0.494 (±0.485)
0.527 (±0.482)
0.127 (±0.268)
0.186 (±0.272)
0.244 (±0.326)
0.362 (±0.428)
Qwen-2.5-32B
0.979 (±0.030)
0.982 (±0.028)
0.603 (±0.471)
0.992 (±0.080)
1.000 (±0.000)
0.800 (±0.400)
0.603 (±0.391)
0.651 (±0.388)
0.826 (±0.341)
Qwen-2.5-3B
0.718 (±0.410)
0.857 (±0.292)
0.374 (±0.434)
0.733 (±0.431)
0.803 (±0.384)
0.453 (±0.451)
0.407 (±0.395)
0.479 (±0.394)
0.603 (±0.441)
Qwen-2.5-72B
0.871 (±0.317)
0.987 (±0.025)
0.614 (±0.471)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.731 (±0.369)
0.811 (±0.329)
0.877 (±0.300)
Qwen-2.0-57B-A14B
0.741 (±0.399)
0.932 (±0.173)
0.222 (±0.370)
0.860 (±0.345)
0.895 (±0.307)
0.630 (±0.438)
0.510 (±0.399)
0.599 (±0.393)
0.673 (±0.424)
Qwen-2.5-7B
0.966 (±0.139)
0.973 (±0.119)
0.329 (±0.411)
0.917 (±0.258)
0.976 (±0.136)
0.586 (±0.459)
0.565 (±0.397)
0.603 (±0.391)
0.739 (±0.394)
Qwen-2.5-Coder-32B
0.937 (±0.219)
0.991 (±0.017)
0.478 (±0.476)
1.000 (±0.000)
1.000 (±0.000)
1.000 (±0.000)
0.814 (±0.325)
0.830 (±0.314)
0.881 (±0.297)
Qwen-2.0-72B
0.964 (±0.040)
0.971 (±0.038)
0.339 (±0.426)
0.950 (±0.199)
1.000 (±0.000)
1.000 (±0.000)
0.630 (±0.380)
0.688 (±0.365)
0.818 (±0.338)
Qwen-2.0-7B
0.566 (±0.452)
0.739 (±0.405)
0.232 (±0.305)
0.799 (±0.397)
0.836 (±0.365)
0.573 (±0.476)
0.298 (±0.359)
0.369 (±0.389)
0.551 (±0.452)
Qwen-3-235B
0.912 (±0.260)
0.993 (±0.017)
0.980 (±0.139)
0.960 (±0.196)
0.993 (±0.086)
1.000 (±0.000)
0.704 (±0.396)
0.813 (±0.337)
0.919 (±0.246)

Leaderboard for RDF Knowledge Graph(KG) related capabilities of Large Language Models(LLMs) as generated with the LLM-KG-Bench framework.

Results for more than 20 RDF related tasks are collected for more than 40 LLMs. The leaderboard contains summarized results:

  • board_combined_scores_short.csv: the most concise summary, listing for each LLM combined scores in the RDF (R) and SPARQL (S) handling categories, estimating read(R) and write(W) capabilities for syntax(syn) and semantic(sem). If a dialogue was tested, the result for the first(1) and best(max) answer is given. board_combined_scores.csv is optimized for automated parsing.
  • board_main_scores.csv contains the summarized stats for the most important scores
  • board_all_stats.csv contains the summarized stats for all scores of all tasks

The full leaderboard with more explanation can be found at https://aksw.github.io/LLM-KG-Bench-Leaderboard/

Downloads last month
-