List your company

Selective Question Answering on TriviaQA

2,849Score (Coverage 14%)LLM-RAG
809.561,339.031,868.52,397.97May 19, 2026

Evaluation results

MethodLinks
LLM-RAGLLM=Qwen3-4B2026.052,8492,5052,6272,7592,8352,8492,8492,8492,849
BalanceRAGLLM=Qwen3-4B2026.052,839—2,5492,7182,8072,8432,8432,8432,843
BalanceRAGLLM=Qwen2.5-7B2026.052,7382,3662,5732,7022,7372,7382,7382,7382,738
LLM-RAGLLM=Qwen3-14B2026.052,7072,5212,6602,7062,7072,7072,7072,7072,707
LLM-RAGLLM=Qwen2.5-7B2026.052,6912,1182,3202,4902,6332,6942,6942,6942,694
BalanceRAGLLM=Llama3.1-8B2026.052,6912,0732,2882,4302,5842,7272,7342,7362,737
BalanceRAGLLM=Qwen3-14B2026.052,6602,4992,6342,6292,6292,6612,6612,6612,661
LLM-RAGLLM=Llama3.1-8B2026.052,5291,8562,0362,1982,3822,6522,7112,7132,713
UCB-Cascaded-CPLLM=Qwen3-14B2026.052,467——2,3102,3622,5382,5332,5252,497
UCB-Cascaded-CPLLM=Qwen2.5-7B2026.052,4421,8201,9922,2842,3772,4992,5312,5342,557
UCB-Cascaded-CPLLM=Llama3.1-8B2026.052,3601,5531,8492,0842,2642,4342,4862,5052,482
LLM-onlyLLM=Qwen2.5-7B2026.052,3471,8561,9882,1482,2552,4222,4732,5222,564
LLM-RAGLLM=Qwen2.5-3B2026.052,2801,4731,6801,9072,1052,4002,4752,5002,501
BalanceRAGLLM=Qwen2.5-3B2026.052,2791,5691,7331,9072,1102,4212,4982,5122,514
UCB-Cascaded-HFDLLM=Qwen2.5-7B2026.052,205—1,7161,8802,0352,3012,3922,4552,502
LLM-onlyLLM=Llama3.1-8B2026.052,1651,5871,7801,9282,0702,2182,2772,3452,402
UCB-Cascaded-CPLLM=Qwen3-4B2026.052,126—1,7461,8342,0162,2402,4632,5332,587
LLM-onlyLLM=Qwen3-4B2026.052,093—1,7641,8711,9952,1592,2332,3012,357
LLM-onlyLLM=Qwen3-14B2026.052,073——1,8361,9642,1762,2592,3142,363
UCB-Cascaded-HFDLLM=Qwen3-14B2026.052,058————2,2152,3592,4102,420
UCB-Cascaded-HFDLLM=Llama3.1-8B2026.051,992—1,3951,6071,8102,1482,2482,3512,400
BalanceRAGLLM=Llama3.2-3B2026.051,9731,3471,6431,7741,8732,0732,1592,2292,290
UCB-Cascaded-CPLLM=Qwen2.5-3B2026.051,952—1,2911,6331,8342,0492,1432,2372,304
UCB-Cascaded-HFDLLM=Qwen3-4B2026.051,902——1,7261,7812,0482,1332,2242,338
UCB-Cascaded-CPLLM=Llama3.2-3B2026.051,8508321,0281,4881,7491,9171,9912,0652,142
LLM-RAGLLM=Llama3.2-3B2026.051,7601,3671,3901,5321,6461,8761,9492,0362,117
UCB-Cascaded-HFDLLM=Qwen2.5-3B2026.051,555————1,7991,9552,0512,152
BalanceRAGLLM=Vicuna-7B-v1.52026.051,5331,2971,3671,4301,4871,5801,6291,6671,694
LLM-RAGLLM=Vicuna-7B-v1.52026.051,5121,2561,3301,4011,4551,5641,6111,6481,675
UCB-Cascaded-CPLLM=Vicuna-7B-v1.52026.051,3549811,1021,2361,3061,3951,4261,4691,504
UCB-Cascaded-HFDLLM=Llama3.2-3B2026.051,328——8209641,6611,7991,8991,990
LLM-onlyLLM=Llama3.2-3B2026.051,2838709881,1161,2141,3711,4431,5131,589
BalanceRAGLLM=Vicuna-13B-v1.52026.051,1689831,0411,0911,1241,2031,2371,2621,274
LLM-RAGLLM=Vicuna-13B-v1.52026.051,1018759399871,0411,1631,2071,2501,270
LLM-onlyLLM=Qwen2.5-3B2026.051,049—7988859711,1371,2271,3161,397
LLM-onlyLLM=Vicuna-7B-v1.52026.051,0497928689359921,1151,1891,2501,294
UCB-Cascaded-HFDLLM=Vicuna-7B-v1.52026.051,0115426037768761,1471,2661,3331,355
LLM-onlyLLM=Vicuna-13B-v1.52026.051,0027758619199611,0431,0811,1101,138
UCB-Cascaded-CPLLM=Vicuna-13B-v1.52026.059987508689259621,0311,0701,1001,127
UCB-Cascaded-HFDLLM=Vicuna-13B-v1.52026.05888—6696948069469891,0351,078