Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
Job has been terminated due to a temporary spike in resource usage and may be restarted later.
Error code: JobManagerCrashedError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
src string | question string | gold string | messages list | pred string | passed bool | top_lp list |
|---|---|---|---|---|---|---|
procedural_aime/diophantine | "Three positive integers x, y, z satisfy x + y + z = 27, x * y = 72, and y * z = 90, with 2 <= x <= (...TRUNCATED) | 2241 | [{"role":"user","content":"Three positive integers x, y, z satisfy x + y + z = 27, x * y = 72, and y(...TRUNCATED) | 241 | false | [{"token":"We","logprob":-0.0420784130692482,"bytes":[87,101],"top_logprobs":[{"token":"We","logprob(...TRUNCATED) |
procedural_aime/polynomial_root_sum | "The cubic equation x^3 + (6)x^2 + (-37)x + (-210) = 0 has three integer roots. Compute the sum of t(...TRUNCATED) | 110 | [{"role":"user","content":"The cubic equation x^3 + (6)x^2 + (-37)x + (-210) = 0 has three integer r(...TRUNCATED) | 3 | false | [{"token":"We","logprob":-0.04062773287296295,"bytes":[87,101],"top_logprobs":[{"token":"We","logpro(...TRUNCATED) |
procedural_aime/factor_chain | "The integer 52003 factors uniquely as a product of four distinct primes. Let p_1 < p_2 < p_3 < p_4 (...TRUNCATED) | 120 | [{"role":"user","content":"The integer 52003 factors uniquely as a product of four distinct primes. (...TRUNCATED) | 120 | true | [{"token":"We","logprob":-0.03961559385061264,"bytes":[87,101],"top_logprobs":[{"token":"We","logpro(...TRUNCATED) |
procedural_aime/lcm_residue | "Find the smallest positive integer x with x mod 5 = 0, x mod 13 = 7, and x mod 9 = 1. The answer is(...TRUNCATED) | 280 | [{"role":"user","content":"Find the smallest positive integer x with x mod 5 = 0, x mod 13 = 7, and (...TRUNCATED) | 280 | true | [{"token":"We","logprob":-0.05224563926458359,"bytes":[87,101],"top_logprobs":[{"token":"We","logpro(...TRUNCATED) |
procedural_aime/iterated_digit | "Define f(n) = 7 * (sum of digits of n) + 3. Starting at n_0 = 456, compute n_5 = f(f(f(f(f(n_0)))))(...TRUNCATED) | 66 | [{"role":"user","content":"Define f(n) = 7 * (sum of digits of n) + 3. Starting at n_0 = 456, comput(...TRUNCATED) | 66 | true | [{"token":"We","logprob":-0.06358783692121506,"bytes":[87,101],"top_logprobs":[{"token":"We","logpro(...TRUNCATED) |
procedural_aime/chained_modular | "Compute the value of ((3^23) * 6 + 14) mod 97, where ^ denotes integer exponentiation. The final an(...TRUNCATED) | 12 | [{"role":"user","content":"Compute the value of ((3^23) * 6 + 14) mod 97, where ^ denotes integer ex(...TRUNCATED) | 273 | false | [{"token":"We","logprob":-0.09161733090877533,"bytes":[87,101],"top_logprobs":[{"token":"We","logpro(...TRUNCATED) |
procedural_aime/factor_chain | "The integer 39767 factors uniquely as a product of four distinct primes. Let p_1 < p_2 < p_3 < p_4 (...TRUNCATED) | 60 | [{"role":"user","content":"The integer 39767 factors uniquely as a product of four distinct primes. (...TRUNCATED) | 60 | true | [{"token":"We","logprob":-0.041340239346027374,"bytes":[87,101],"top_logprobs":[{"token":"We","logpr(...TRUNCATED) |
procedural_aime/polynomial_root_sum | "The cubic equation x^3 + (12)x^2 + (32)x + (0) = 0 has three integer roots. Compute the sum of the (...TRUNCATED) | 80 | [{"role":"user","content":"The cubic equation x^3 + (12)x^2 + (32)x + (0) = 0 has three integer root(...TRUNCATED) | 0 | false | [{"token":"We","logprob":-0.049532800912857056,"bytes":[87,101],"top_logprobs":[{"token":"We","logpr(...TRUNCATED) |
procedural_aime/lcm_residue | "Find the smallest positive integer x with x mod 7 = 5, x mod 8 = 5, and x mod 5 = 3. The answer is (...TRUNCATED) | 173 | [{"role":"user","content":"Find the smallest positive integer x with x mod 7 = 5, x mod 8 = 5, and x(...TRUNCATED) | 3 | false | [{"token":"We","logprob":-0.03868890926241875,"bytes":[87,101],"top_logprobs":[{"token":"We","logpro(...TRUNCATED) |
procedural_aime/coordinate_geometry | "Count the number of lattice points (x, y) with 1 <= x <= 17 and 1 <= y <= 13 such that gcd(x, y) = (...TRUNCATED) | 148 | [{"role":"user","content":"Count the number of lattice points (x, y) with 1 <= x <= 17 and 1 <= y <=(...TRUNCATED) | -1 | false | [{"token":"We","logprob":-0.12272240221500397,"bytes":[87,101],"top_logprobs":[{"token":"We","logpro(...TRUNCATED) |
End of preview.
Dataset Card for Kimi K 2.6 Rollout
Dataset Description
5000 items -> math_dataset.jsonl
============================================================
SUMMARY
============================================================
📊 Overall: 1157/5000 passed (23.14%)
Kind Passed Total Rate
------------------------------------------------------------
arithmetic_series 24 82 29.27%
bakery_orders 40 292 13.70%
coin_change 33 84 39.29%
construction 10 291 3.44%
digit_sum 63 84 75.00%
factorial_mod 39 83 46.99%
fundraiser 9 292 3.08%
garden_orchard 9 292 3.08%
gcd_lcm 21 84 25.00%
geometric_series 25 84 29.76%
library_books 67 292 22.95%
mixture 22 84 26.19%
modular_linear 31 83 37.35%
percentage 4 82 4.88%
pets_animals 4 292 1.37%
polynomial_eval 38 84 45.24%
probability_count 41 83 49.40%
proportion 31 82 37.80%
rate_distance 41 84 48.81%
recipe_scale 102 291 35.05%
school_classroom 118 291 40.55%
set_intersect 37 83 44.58%
shopping_budget 6 292 2.05%
simultaneous 30 84 35.71%
sports_tournament 105 291 36.08%
time_arithmetic 21 84 25.00%
travel_distance 88 292 30.14%
triangle_area 55 84 65.48%
trip_planning 3 292 1.03%
unit_conversion 40 82 48.78%
============================================================
3000 items -> ifeval_dataset.jsonl
============================================================
EVALUATION SUMMARY
============================================================
📊 Overall: 1373/3000 passed (45.77%)
Kind Passed Total Rate
------------------------------------------------------------
all_lowercase 78 93 83.87%
all_uppercase 77 92 83.70%
bullet_list 30 92 32.61%
compound3:all_lowercase+i 15 75 20.00%
compound3:bullet_list+inc 4 87 4.60%
compound3:bullet_list+min 0 80 0.00%
compound3:exact_sentences 13 70 18.57%
compound3:max_words+no_co 17 72 23.61%
compound3:min_words+all_l 0 76 0.00%
compound3:min_words+forbi 10 75 13.33%
compound3:min_words+inclu 3 73 4.11%
compound3:title_format+ex 1 63 1.59%
compound3:title_format+ex 14 79 17.72%
compound:all_lowercase+en 0 71 0.00%
compound:all_lowercase+in 14 89 15.73%
compound:bullet_list+incl 26 86 30.23%
compound:bullet_list+min_ 53 74 71.62%
compound:exact_sentences+ 20 85 23.53%
compound:exact_sentences+ 40 66 60.61%
compound:forbid_keyword+e 78 85 91.76%
compound:forbid_keyword+m 38 77 49.35%
compound:max_words+no_com 28 61 45.90%
compound:min_words+ends_w 37 81 45.68%
compound:min_words+includ 66 76 86.84%
compound:title_format+end 70 104 67.31%
compound:title_format+exa 68 95 71.58%
ends_with_phrase 43 92 46.74%
exact_sentences 77 92 83.70%
exact_words 16 93 17.20%
forbid_keyword 62 92 67.39%
include_keyword 73 92 79.35%
json_format 34 92 36.96%
max_words 34 93 36.56%
min_words 82 93 88.17%
no_comma 61 92 66.30%
title_format 91 92 98.91%
============================================================
5000 items -> aime_dataset.jsonl
============================================================
EVALUATION SUMMARY
============================================================
📊 Overall: 2821/5000 passed (56.42%)
Kind Passed Total Rate
------------------------------------------------------------
chained_modular 385 714 53.92%
coordinate_geometry 84 714 11.76%
diophantine 183 715 25.59%
factor_chain 563 715 78.74%
iterated_digit 481 714 67.37%
lcm_residue 562 714 78.71%
polynomial_root_sum 563 714 78.85%
============================================================
TOOL_USE_DATASET 6000
============================================================
EVALUATION SUMMARY
============================================================
📊 Overall: 915/6000 passed (15.25%)
Kind Passed Total Rate
------------------------------------------------------------
arithmetic_series 25 101 24.75%
bakery_orders 50 351 14.25%
coin_change 18 102 17.65%
construction 20 350 5.71%
digit_sum 20 102 19.61%
factorial_mod 28 102 27.45%
fundraiser 11 349 3.15%
garden_orchard 14 348 4.02%
gcd_lcm 25 100 25.00%
geometric_series 26 99 26.26%
library_books 62 350 17.71%
mixture 29 99 29.29%
modular_linear 25 100 25.00%
percentage 5 100 5.00%
pets_animals 4 349 1.15%
polynomial_eval 27 100 27.00%
probability_count 34 100 34.00%
proportion 15 98 15.31%
rate_distance 24 99 24.24%
recipe_scale 97 352 27.57%
school_classroom 61 348 17.53%
set_intersect 26 98 26.53%
shopping_budget 5 352 1.42%
simultaneous 15 99 15.15%
sports_tournament 85 352 24.14%
time_arithmetic 25 102 24.51%
travel_distance 93 348 26.74%
triangle_area 19 98 19.39%
trip_planning 6 351 1.71%
unit_conversion 21 102 20.59%
============================================================
Generated 6000 items -> robustness_dataset.jsonl
============================================================
EVALUATION SUMMARY
============================================================
📊 Overall: 1291/6000 passed (21.52%)
Kind Passed Total Rate
------------------------------------------------------------
arithmetic_series 24 102 23.53%
bakery_orders 43 352 12.22%
coin_change 40 102 39.22%
construction 16 348 4.60%
digit_sum 63 102 61.76%
factorial_mod 40 102 39.22%
fundraiser 5 350 1.43%
garden_orchard 22 350 6.29%
gcd_lcm 22 100 22.00%
geometric_series 20 100 20.00%
library_books 53 352 15.06%
mixture 27 102 26.47%
modular_linear 48 100 48.00%
percentage 4 100 4.00%
pets_animals 8 346 2.31%
polynomial_eval 49 100 49.00%
probability_count 51 100 51.00%
proportion 38 100 38.00%
rate_distance 39 96 40.62%
recipe_scale 124 352 35.23%
school_classroom 154 348 44.25%
set_intersect 45 98 45.92%
shopping_budget 5 352 1.42%
simultaneous 32 98 32.65%
sports_tournament 101 352 28.69%
time_arithmetic 22 102 21.57%
travel_distance 95 346 27.46%
triangle_area 56 96 58.33%
trip_planning 1 352 0.28%
unit_conversion 44 100 44.00%
============================================================
[gen] Generated 10000 items -> multi_doc_dataset.jsonl
============================================================
EVALUATION SUMMARY
============================================================
📊 Overall: 589/10000 passed (5.89%)
Kind Passed Total Rate
------------------------------------------------------------
compare 68 1667 4.08%
difference 16 1667 0.96%
difference_three 6 1666 0.36%
ratio 483 1667 28.97%
sum 14 1667 0.84%
sum_three 2 1666 0.12%
============================================================
[gen] Generated 6000 items -> reasoning_dataset.jsonl
============================================================
EVALUATION SUMMARY
============================================================
📊 Overall: 673/6000 passed (11.22%)
Kind Passed Total Rate
------------------------------------------------------------
analogy_letter 52 301 17.28%
boolean_eval 82 299 27.42%
constraint_sat 59 601 9.82%
date_arithmetic 36 601 5.99%
deduction 32 301 10.63%
graph_connectivity 33 600 5.50%
graph_shortest_path 83 599 13.86%
navigate_steps 4 598 0.67%
odd_one_out 100 301 33.22%
ordering 79 301 26.25%
sequence_next 41 297 13.80%
tracking_objects 38 601 6.32%
web_of_lies 34 600 5.67%
============================================================
[gen] Generated 3000 items -> code_dataset.jsonl
============================================================
EVALUATION SUMMARY
============================================================
📊 Overall: 1687/3000 passed (56.23%)
Kind Passed Total Rate
------------------------------------------------------------
aggregate_list 32 38 84.21%
binary_search_first 95 128 74.22%
caesar_cipher 79 128 61.72%
closest_pair_sum 37 128 28.91%
coin_change_min 20 129 15.50%
compress_string 18 128 14.06%
digit_sum 29 38 76.32%
evaluate_postfix 57 129 44.19%
flatten_nested_list 66 129 51.16%
group_anagrams 69 129 53.49%
kth_largest 61 129 47.29%
longest_no_repeat 85 128 66.41%
longest_palindromic_subst 33 128 25.78%
merge_intervals 33 128 25.78%
most_freq_elem 114 129 88.37%
nested_paren_groups 62 129 48.06%
pair_count 28 38 73.68%
rolling_max 88 129 68.22%
roman_to_int 91 128 71.09%
run_length 24 37 64.86%
run_length_decode 77 129 59.69%
sliding_window_min 94 129 72.87%
string_predicate 30 38 78.95%
string_transform 30 36 83.33%
transform_list 32 38 84.21%
two_sum_pairs 92 128 71.88%
unique_paths_grid 89 129 68.99%
validate_brackets 94 129 72.87%
window_sum 28 37 75.68%
============================================================
Usage
You can load the Kimi K 2.6 dataset using the datasets library:
from datasets import load_dataset
dataset = load_dataset("kimi_k_2.6")
- Downloads last month
- 5