id
stringclasses
252 values
prompt
stringlengths
2.5k
15.1k
gt_is_correct
bool
2 classes
gt_token_count
int64
1
29.5k
source_model
stringclasses
19 values
retrieved_anchors
listlengths
5
5
mmlu_pro_006227
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,244
tngtech/deepseek-r1t2-chimera
[ "anchor203", "anchor166", "anchor194", "anchor207", "anchor007" ]
Forum-283-1
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
2,828
tngtech/deepseek-r1t2-chimera:free
[ "anchor166", "anchor209", "anchor165", "anchor212", "anchor216" ]
rbench_en_000586
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,738
tngtech/deepseek-r1t2-chimera
[ "anchor070", "anchor187", "anchor192", "anchor231", "anchor212" ]
2010-2022_Geography_MCQs_33
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
601
tngtech/deepseek-r1t2-chimera:free
[ "anchor037", "anchor014", "anchor038", "anchor049", "anchor013" ]
mmlu_pro_000234
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
1,119
tngtech/deepseek-r1t2-chimera
[ "anchor060", "anchor132", "anchor054", "anchor134", "anchor049" ]
2010-2022_Math_II_MCQs_173
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
940
tngtech/deepseek-r1t2-chimera:free
[ "anchor230", "anchor245", "anchor002", "anchor239", "anchor003" ]
rbench_en_000456
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
9,259
tngtech/deepseek-r1t2-chimera
[ "anchor076", "anchor243", "anchor242", "anchor230", "anchor073" ]
mmlu_pro_007020
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
540
tngtech/deepseek-r1t2-chimera
[ "anchor081", "anchor129", "anchor038", "anchor181", "anchor156" ]
2010-2022_Chemistry_MCQs_78
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
868
tngtech/deepseek-r1t2-chimera:free
[ "anchor170", "anchor182", "anchor176", "anchor178", "anchor019" ]
gpqa_000385
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
853
tngtech/deepseek-r1t2-chimera
[ "anchor123", "anchor206", "anchor054", "anchor057", "anchor217" ]
rbench_en_001009
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
4,299
tngtech/deepseek-r1t2-chimera
[ "anchor027", "anchor137", "anchor138", "anchor158", "anchor167" ]
mmlu_pro_009646
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
1,874
tngtech/deepseek-r1t2-chimera
[ "anchor160", "anchor027", "anchor140", "anchor029", "anchor137" ]
mmlu_pro_010863
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,115
tngtech/deepseek-r1t2-chimera
[ "anchor133", "anchor229", "anchor232", "anchor226", "anchor038" ]
2010-2022_History_MCQs_224
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
680
tngtech/deepseek-r1t2-chimera:free
[ "anchor011", "anchor229", "anchor014", "anchor013", "anchor039" ]
mmlu_pro_010458
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
766
tngtech/deepseek-r1t2-chimera
[ "anchor074", "anchor114", "anchor048", "anchor196", "anchor119" ]
2010-2022_History_MCQs_195
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
597
tngtech/deepseek-r1t2-chimera:free
[ "anchor039", "anchor013", "anchor045", "anchor103", "anchor014" ]
2010-2022_History_MCQs_265
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
608
tngtech/deepseek-r1t2-chimera:free
[ "anchor047", "anchor038", "anchor039", "anchor013", "anchor009" ]
mmlu_pro_011674
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
5,741
tngtech/deepseek-r1t2-chimera
[ "anchor069", "anchor068", "anchor137", "anchor220", "anchor017" ]
2012-2022_English_Cloze_Test_16
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
605
tngtech/deepseek-r1t2-chimera:free
[ "anchor037", "anchor123", "anchor010", "anchor127", "anchor046" ]
2010-2022_Math_I_MCQs_202
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
725
tngtech/deepseek-r1t2-chimera:free
[ "anchor228", "anchor240", "anchor239", "anchor248", "anchor246" ]
Forum-158-1
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
6,331
tngtech/deepseek-r1t2-chimera:free
[ "anchor242", "anchor003", "anchor119", "anchor021", "anchor233" ]
mmlu_pro_001628
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
996
tngtech/deepseek-r1t2-chimera
[ "anchor061", "anchor024", "anchor032", "anchor108", "anchor062" ]
mmlu_pro_002489
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
594
tngtech/deepseek-r1t2-chimera
[ "anchor134", "anchor132", "anchor130", "anchor122", "anchor039" ]
2010-2022_Math_I_MCQs_41
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
479
tngtech/deepseek-r1t2-chimera:free
[ "anchor245", "anchor002", "anchor039", "anchor003", "anchor242" ]
2010-2013_English_MCQs_52
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
502
tngtech/deepseek-r1t2-chimera:free
[ "anchor039", "anchor014", "anchor013", "anchor023", "anchor038" ]
Forum-27-1
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,193
tngtech/deepseek-r1t2-chimera:free
[ "anchor021", "anchor073", "anchor145", "anchor048", "anchor039" ]
rbench_en_000891
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,259
tngtech/deepseek-r1t2-chimera
[ "anchor027", "anchor245", "anchor172", "anchor120", "anchor074" ]
mmlu_pro_007101
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
720
tngtech/deepseek-r1t2-chimera
[ "anchor024", "anchor022", "anchor080", "anchor085", "anchor025" ]
gpqa_000024
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
6,295
tngtech/deepseek-r1t2-chimera
[ "anchor195", "anchor198", "anchor196", "anchor027", "anchor078" ]
mmlu_pro_001140
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,073
tngtech/deepseek-r1t2-chimera
[ "anchor110", "anchor032", "anchor214", "anchor035", "anchor208" ]
rbench_en_000528
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
4,911
tngtech/deepseek-r1t2-chimera
[ "anchor005", "anchor116", "anchor222", "anchor232", "anchor223" ]
2010-2022_Political_Science_MCQs_272
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
507
tngtech/deepseek-r1t2-chimera:free
[ "anchor039", "anchor245", "anchor047", "anchor038", "anchor013" ]
rbench_en_000661
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
2,484
tngtech/deepseek-r1t2-chimera
[ "anchor137", "anchor087", "anchor088", "anchor027", "anchor140" ]
2010-2022_Biology_MCQs_65
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
512
tngtech/deepseek-r1t2-chimera:free
[ "anchor204", "anchor055", "anchor039", "anchor047", "anchor048" ]
rbench_en_000995
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
9,225
tngtech/deepseek-r1t2-chimera
[ "anchor137", "anchor029", "anchor160", "anchor135", "anchor190" ]
rbench_en_000203
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
7,430
tngtech/deepseek-r1t2-chimera
[ "anchor087", "anchor139", "anchor028", "anchor141", "anchor027" ]
gpqa_000124
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
2,530
tngtech/deepseek-r1t2-chimera
[ "anchor160", "anchor137", "anchor136", "anchor135", "anchor092" ]
gpqa_000010
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
742
tngtech/deepseek-r1t2-chimera
[ "anchor159", "anchor206", "anchor212", "anchor162", "anchor166" ]
2010-2022_History_MCQs_201
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
700
tngtech/deepseek-r1t2-chimera:free
[ "anchor039", "anchor038", "anchor047", "anchor048", "anchor049" ]
2010-2022_Biology_MCQs_9
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
1,455
tngtech/deepseek-r1t2-chimera:free
[ "anchor204", "anchor166", "anchor019", "anchor039", "anchor047" ]
2010-2022_Chemistry_MCQs_102
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
686
tngtech/deepseek-r1t2-chimera:free
[ "anchor220", "anchor019", "anchor018", "anchor183", "anchor176" ]
mmlu_pro_011882
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
6,013
tngtech/deepseek-r1t2-chimera
[ "anchor137", "anchor028", "anchor017", "anchor143", "anchor155" ]
rbench_en_000901
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,294
tngtech/deepseek-r1t2-chimera
[ "anchor120", "anchor245", "anchor003", "anchor192", "anchor209" ]
2010-2022_History_MCQs_152
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
463
tngtech/deepseek-r1t2-chimera:free
[ "anchor047", "anchor245", "anchor014", "anchor013", "anchor039" ]
rbench_en_000531
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
3,834
tngtech/deepseek-r1t2-chimera
[ "anchor250", "anchor116", "anchor222", "anchor242", "anchor233" ]
2010-2022_History_MCQs_267
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
609
tngtech/deepseek-r1t2-chimera:free
[ "anchor011", "anchor047", "anchor039", "anchor013", "anchor045" ]
2010-2022_Math_II_MCQs_87
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
1,049
tngtech/deepseek-r1t2-chimera:free
[ "anchor004", "anchor237", "anchor233", "anchor249", "anchor227" ]
gpqa_000047
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
false
789
tngtech/deepseek-r1t2-chimera
[ "anchor163", "anchor057", "anchor162", "anchor055", "anchor008" ]
gpqa_000225
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
3,315
tngtech/deepseek-r1t2-chimera
[ "anchor195", "anchor198", "anchor220", "anchor196", "anchor137" ]
Forum-352-1
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
4,745
tngtech/deepseek-r1t2-chimera:free
[ "anchor160", "anchor195", "anchor198", "anchor151", "anchor197" ]
2010-2022_Biology_MCQs_53
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
352
tngtech/deepseek-r1t2-chimera:free
[ "anchor055", "anchor048", "anchor208", "anchor057", "anchor010" ]
2010-2022_History_MCQs_75
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model tngtech/deeps...
true
578
tngtech/deepseek-r1t2-chimera:free
[ "anchor038", "anchor039", "anchor245", "anchor047", "anchor021" ]
rbench_en_000042
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
1,615
google/gemma-3-12b-it
[ "anchor090", "anchor092", "anchor029", "anchor158", "anchor138" ]
mmlu_pro_011758
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
646
google/gemma-3-12b-it
[ "anchor220", "anchor086", "anchor017", "anchor190", "anchor137" ]
2010-2022_Math_I_MCQs_91
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
326
google/gemma-3-12b-it
[ "anchor223", "anchor245", "anchor241", "anchor047", "anchor039" ]
gpqa_000155
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
57
google/gemma-3-12b-it
[ "anchor180", "anchor181", "anchor170", "anchor175", "anchor176" ]
2010-2022_History_MCQs_2
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
280
google/gemma-3-12b-it
[ "anchor039", "anchor013", "anchor047", "anchor038", "anchor049" ]
2010-2022_Biology_MCQs_111
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
457
google/gemma-3-12b-it
[ "anchor208", "anchor213", "anchor209", "anchor211", "anchor039" ]
rbench_en_000492
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
1,963
google/gemma-3-12b-it
[ "anchor241", "anchor250", "anchor243", "anchor224", "anchor115" ]
2010-2022_Chinese_Lang_and_Usage_MCQs_40
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
227
google/gemma-3-12b-it
[ "anchor014", "anchor039", "anchor013", "anchor011", "anchor245" ]
2010-2022_History_MCQs_233
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
302
google/gemma-3-12b-it
[ "anchor039", "anchor245", "anchor002", "anchor047", "anchor122" ]
rbench_en_000572
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
1,449
google/gemma-3-12b-it
[ "anchor116", "anchor222", "anchor239", "anchor242", "anchor241" ]
mmlu_pro_001559
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
5
google/gemma-3-12b-it
[ "anchor113", "anchor107", "anchor111", "anchor108", "anchor040" ]
2010-2022_English_Reading_Comp_99
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
191
google/gemma-3-12b-it
[ "anchor037", "anchor148", "anchor166", "anchor207", "anchor040" ]
gpqa_000265
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
32
google/gemma-3-12b-it
[ "anchor182", "anchor185", "anchor176", "anchor181", "anchor186" ]
mmlu_pro_005256
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
5
google/gemma-3-12b-it
[ "anchor112", "anchor127", "anchor053", "anchor045", "anchor103" ]
rbench_en_000890
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
1,236
google/gemma-3-12b-it
[ "anchor229", "anchor201", "anchor116", "anchor248", "anchor079" ]
2010-2022_Political_Science_MCQs_16
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
305
google/gemma-3-12b-it
[ "anchor039", "anchor047", "anchor034", "anchor038", "anchor132" ]
rbench_en_001084
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
16,384
google/gemma-3-12b-it
[ "anchor201", "anchor219", "anchor241", "anchor005", "anchor227" ]
2010-2022_Math_II_MCQs_91
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
122
google/gemma-3-12b-it
[ "anchor245", "anchor241", "anchor003", "anchor013", "anchor229" ]
mmlu_pro_001593
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
248
google/gemma-3-12b-it
[ "anchor110", "anchor032", "anchor108", "anchor112", "anchor124" ]
rbench_en_000457
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
1,211
google/gemma-3-12b-it
[ "anchor230", "anchor115", "anchor245", "anchor201", "anchor241" ]
2010-2022_Political_Science_MCQs_209
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
332
google/gemma-3-12b-it
[ "anchor047", "anchor039", "anchor038", "anchor245", "anchor037" ]
2010-2022_Political_Science_MCQs_219
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
335
google/gemma-3-12b-it
[ "anchor039", "anchor038", "anchor013", "anchor047", "anchor049" ]
gpqa_000255
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
9
google/gemma-3-12b-it
[ "anchor187", "anchor186", "anchor169", "anchor170", "anchor178" ]
2010-2022_History_MCQs_6
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
323
google/gemma-3-12b-it
[ "anchor039", "anchor002", "anchor036", "anchor038", "anchor047" ]
mmlu_pro_001213
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
5
google/gemma-3-12b-it
[ "anchor058", "anchor108", "anchor112", "anchor063", "anchor035" ]
mmlu_pro_001741
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
238
google/gemma-3-12b-it
[ "anchor108", "anchor032", "anchor061", "anchor058", "anchor229" ]
2010-2022_English_Reading_Comp_10
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
332
google/gemma-3-12b-it
[ "anchor023", "anchor014", "anchor127", "anchor039", "anchor245" ]
2010-2022_History_MCQs_237
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
394
google/gemma-3-12b-it
[ "anchor039", "anchor049", "anchor245", "anchor013", "anchor014" ]
2010-2022_Chinese_Modern_Lit_16
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
264
google/gemma-3-12b-it
[ "anchor047", "anchor039", "anchor038", "anchor034", "anchor049" ]
mmlu_pro_007068
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
5
google/gemma-3-12b-it
[ "anchor022", "anchor023", "anchor082", "anchor047", "anchor024" ]
mmlu_pro_010120
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
2,066
google/gemma-3-12b-it
[ "anchor137", "anchor087", "anchor091", "anchor092", "anchor029" ]
mmlu_pro_009046
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
557
google/gemma-3-12b-it
[ "anchor090", "anchor092", "anchor027", "anchor087", "anchor138" ]
mmlu_pro_011872
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
1,079
google/gemma-3-12b-it
[ "anchor137", "anchor068", "anchor087", "anchor017", "anchor139" ]
mmlu_pro_006935
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
5
google/gemma-3-12b-it
[ "anchor022", "anchor124", "anchor082", "anchor025", "anchor024" ]
2010-2022_History_MCQs_226
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
359
google/gemma-3-12b-it
[ "anchor050", "anchor038", "anchor011", "anchor013", "anchor021" ]
2010-2022_History_MCQs_206
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
373
google/gemma-3-12b-it
[ "anchor038", "anchor039", "anchor014", "anchor013", "anchor047" ]
2010-2022_Political_Science_MCQs_21
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
298
google/gemma-3-12b-it
[ "anchor038", "anchor039", "anchor049", "anchor245", "anchor013" ]
mmlu_pro_011200
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
1,433
google/gemma-3-12b-it
[ "anchor089", "anchor069", "anchor087", "anchor137", "anchor027" ]
mmlu_pro_001816
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
518
google/gemma-3-12b-it
[ "anchor108", "anchor032", "anchor110", "anchor129", "anchor107" ]
gpqa_000209
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
15
google/gemma-3-12b-it
[ "anchor152", "anchor200", "anchor195", "anchor158", "anchor137" ]
mmlu_pro_010110
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
807
google/gemma-3-12b-it
[ "anchor067", "anchor142", "anchor167", "anchor197", "anchor171" ]
rbench_en_000784
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
1,999
google/gemma-3-12b-it
[ "anchor043", "anchor120", "anchor245", "anchor209", "anchor164" ]
rbench_en_000741
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
1,063
google/gemma-3-12b-it
[ "anchor072", "anchor043", "anchor061", "anchor164", "anchor181" ]
2010-2022_Math_II_MCQs_32
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
539
google/gemma-3-12b-it
[ "anchor230", "anchor237", "anchor004", "anchor241", "anchor245" ]
rbench_en_000919
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
1,121
google/gemma-3-12b-it
[ "anchor079", "anchor078", "anchor154", "anchor115", "anchor241" ]
2012-2022_English_Cloze_Test_0
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
14
google/gemma-3-12b-it
[ "anchor127", "anchor204", "anchor003", "anchor145", "anchor094" ]
rbench_en_000040
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
true
1,060
google/gemma-3-12b-it
[ "anchor027", "anchor029", "anchor092", "anchor090", "anchor066" ]
Forum-228-1
### Task You are a performance prediction expert. Given a target question, 5 anchor questions with their performance results, and a target AI model, predict how the model will perform on the target question, specifically the output length and correctness after related reasoning analysis. ### Target Model google/gemma-...
false
6
google/gemma-3-12b-it
[ "anchor161", "anchor057", "anchor074", "anchor048", "anchor213" ]